\n \n\n\n

Fashion Show Attendance Trends: 2005-2020

\n \n \n\n \n\n "}}},{"rowIdx":2,"cells":{"id":{"kind":"string","value":"PlotlyChartPipeline_density_1-98"},"image":{"kind":"image","value":{"src":"https://datasets-server.huggingface.co/assets/allenai/CoSyn-400K/--/86e46e1fd5e754d056169f0fb38f06c6997ff7de/--/chart/train/2/image/image.jpg?Expires=1741922929&Signature=w62IbP2c7ya3nsTDTRjtj8DJAswpQREhtIqGWOwvWy2fs7VS0f3fyAAaltYJ6vbjyaLtrYu5qTsH~9l3ZPF4NLAPG8OUYvyfzNaisCeTSN6MpzdOm1PUZGFHAuImUIt~ldnu8McC255YaPRKd8FOCgqR1r1143M4-CEhrBCsrDEfo~kO1j6r68B9SArZPlKkNcXgS8tUA~qFNum6l69vZb3Lxw2HuUirYboUQ3SBXN-L~T3xKzdMZZLZN7HYZe2Q6GBsu4un55pzHZg3ycli05ETueK1nrgxgtagKLX9p0jkXWVbxscD1dz207HYD3SlI3VBFjNJzL2toFHwGhw8jQ__&Key-Pair-Id=K3EI6M078Z3AC3","height":1440,"width":2160},"inferredFromUrl":false},"qa_pairs":{"kind":"list like","value":{"question":["What is the highest market share category?","How many categories have a market share above 10%?","Which category has the smallest market share?","What is the combined market share of the categories ranked second and third?","Is the market share of Holographic Displays greater than that of Brain-Computer Interfaces?","What is the median market share value across all categories?","Which category has roughly half the market share of AR Glasses?","Are there more categories with market share under 5% or above 5%?","What is the range of the market shares displayed in the plot?","How does the market share of Brain-Computer Interfaces compare to that of Smart Contact Lenses?"],"explanation":["The figure reveals that the highest market share percentage is associated with Smartwatches. This observation is based on the fact that the extreme right of the density plot, where the highest x-values are located, corresponds to the 'Smartwatches' category.","Looking at the horizontal density plot, we identify categories with their values extending beyond the 10% mark on the x-axis. These categories are 'Smartwatches' and 'AR Glasses'.","The lowest point on the x-axis of the density plot is associated with the 'Neuralink-style Implants' category, which is positioned farthest left.","The second and third largest market shares are AR Glasses (15.2%) and Smart Clothing (9.7%), respectively. Summing these yields 15.2 + 9.7.","By comparing their positions on the x-axis, Brain-Computer Interfaces (5.3%) have a higher market share than Holographic Displays (3.8%).","Listing the values in ascending order: 0.5, 0.8, 1.2, 1.9, 2.1, 3.8, 5.3, 9.7, 15.2, 28.5, the median is the average of two middle values: (2.1 + 3.8)/2 = 2.95.","Identifying the categories by their values, approximately half of AR Glasses' market share (15.2/2 = 7.6) closely matches Smart Clothing's 9.7%.","Counting the categories, market shares under 5% include Neuralink-style Implants, AI-Enhanced Exoskeletons, Quantum Computing Wearables, Smart Contact Lenses, Implantable Chips, and Holographic Displays. While market shares above 5% include Brain-Computer Interfaces, Smart Clothing, AR Glasses, and Smartwatches. The numbers are 6 versus 4.","The difference between the maximum and minimum market share values provides the range: 28.5% (Smartwatches) - 0.5% (Neuralink-style Implants).","Analyzing the x-axis positions reveals that Brain-Computer Interfaces (5.3%) have a higher market share compared to Smart Contact Lenses (1.9%)."],"answer":["Smartwatches","2","Neuralink-style Implants","24.9%","No","2.95%","Smart Clothing","Under 5%","28.0%","Higher"]},"string":"{\n \"question\": [\n \"What is the highest market share category?\",\n \"How many categories have a market share above 10%?\",\n \"Which category has the smallest market share?\",\n \"What is the combined market share of the categories ranked second and third?\",\n \"Is the market share of Holographic Displays greater than that of Brain-Computer Interfaces?\",\n \"What is the median market share value across all categories?\",\n \"Which category has roughly half the market share of AR Glasses?\",\n \"Are there more categories with market share under 5% or above 5%?\",\n \"What is the range of the market shares displayed in the plot?\",\n \"How does the market share of Brain-Computer Interfaces compare to that of Smart Contact Lenses?\"\n ],\n \"explanation\": [\n \"The figure reveals that the highest market share percentage is associated with Smartwatches. This observation is based on the fact that the extreme right of the density plot, where the highest x-values are located, corresponds to the 'Smartwatches' category.\",\n \"Looking at the horizontal density plot, we identify categories with their values extending beyond the 10% mark on the x-axis. These categories are 'Smartwatches' and 'AR Glasses'.\",\n \"The lowest point on the x-axis of the density plot is associated with the 'Neuralink-style Implants' category, which is positioned farthest left.\",\n \"The second and third largest market shares are AR Glasses (15.2%) and Smart Clothing (9.7%), respectively. Summing these yields 15.2 + 9.7.\",\n \"By comparing their positions on the x-axis, Brain-Computer Interfaces (5.3%) have a higher market share than Holographic Displays (3.8%).\",\n \"Listing the values in ascending order: 0.5, 0.8, 1.2, 1.9, 2.1, 3.8, 5.3, 9.7, 15.2, 28.5, the median is the average of two middle values: (2.1 + 3.8)/2 = 2.95.\",\n \"Identifying the categories by their values, approximately half of AR Glasses' market share (15.2/2 = 7.6) closely matches Smart Clothing's 9.7%.\",\n \"Counting the categories, market shares under 5% include Neuralink-style Implants, AI-Enhanced Exoskeletons, Quantum Computing Wearables, Smart Contact Lenses, Implantable Chips, and Holographic Displays. While market shares above 5% include Brain-Computer Interfaces, Smart Clothing, AR Glasses, and Smartwatches. The numbers are 6 versus 4.\",\n \"The difference between the maximum and minimum market share values provides the range: 28.5% (Smartwatches) - 0.5% (Neuralink-style Implants).\",\n \"Analyzing the x-axis positions reveals that Brain-Computer Interfaces (5.3%) have a higher market share compared to Smart Contact Lenses (1.9%).\"\n ],\n \"answer\": [\n \"Smartwatches\",\n \"2\",\n \"Neuralink-style Implants\",\n \"24.9%\",\n \"No\",\n \"2.95%\",\n \"Smart Clothing\",\n \"Under 5%\",\n \"28.0%\",\n \"Higher\"\n ]\n}"},"metadata":{"kind":"string","value":"{\n \"figure_type\": \"horizontal density plot\",\n \"persona\": \"A technology visionary predicting the future of personal computing, exploring concepts like quantum computing and artificial intelligence in PCs\",\n \"topic\": \"Market share distribution of emerging personal computing form factors (e.g., wearables, implantables, holograms)\"\n}"},"data":{"kind":"string","value":"Category,Market Share\nSmartwatches,28.5\nAR Glasses,15.2\nSmart Clothing,9.7\nBrain-Computer Interfaces,5.3\nHolographic Displays,3.8\nImplantable Chips,2.1\nSmart Contact Lenses,1.9\nQuantum Computing Wearables,1.2\nAI-Enhanced Exoskeletons,0.8\nNeuralink-style Implants,0.5"},"code":{"kind":"string","value":"import plotly.express as px\nimport plotly.graph_objects as go\nfrom io import BytesIO\nfrom PIL import Image\nimport numpy as np\n\ndef generate_plot(data):\n # Sort data by Market Share in descending order\n data = data.sort_values('Market Share', ascending=True)\n\n # Create horizontal density plot\n fig = go.Figure()\n \n fig.add_trace(go.Violin(\n y=data['Category'],\n x=data['Market Share'],\n orientation='h',\n side='positive',\n width=2,\n points=False,\n line_color='rgba(70, 130, 180, 0.7)',\n fillcolor='rgba(135, 206, 235, 0.5)',\n marker=dict(size=10, color='rgba(25, 25, 112, 0.7)')\n ))\n\n # Add markers for actual data points\n fig.add_trace(go.Scatter(\n x=data['Market Share'],\n y=data['Category'],\n mode='markers',\n marker=dict(size=12, color='rgba(25, 25, 112, 0.9)', symbol='diamond'),\n name='Actual Values'\n ))\n\n # Customize layout\n fig.update_layout(\n title={\n 'text': 'Market Share Distribution of Emerging Personal Computing Form Factors',\n 'font': dict(size=24, family='Arial Black', color='#333333'),\n 'y': 0.95,\n 'x': 0.5,\n 'xanchor': 'center',\n 'yanchor': 'top'\n },\n xaxis_title={\n 'text': 'Market Share (%)',\n 'font': dict(size=18, family='Arial', color='#333333')\n },\n yaxis_title={\n 'text': 'Category',\n 'font': dict(size=18, family='Arial', color='#333333')\n },\n xaxis=dict(\n tickfont=dict(size=14, family='Arial'),\n tickmode='linear',\n tick0=0,\n dtick=5,\n range=[0, max(data['Market Share']) * 1.1]\n ),\n yaxis=dict(\n tickfont=dict(size=14, family='Arial'),\n categoryorder='total ascending'\n ),\n height=800,\n width=1200,\n plot_bgcolor='rgba(240, 240, 240, 0.5)',\n paper_bgcolor='white',\n margin=dict(l=200, r=50, t=100, b=100),\n showlegend=False\n )\n\n # Add a futuristic touch with a gradient background\n fig.add_layout_image(\n dict(\n source=\"https://images.plot.ly/language-icons/api-home/python-logo.png\",\n xref=\"paper\", yref=\"paper\",\n x=1, y=1.05,\n sizex=0.2, sizey=0.2,\n xanchor=\"right\", yanchor=\"bottom\",\n opacity=0.8,\n layer=\"below\"\n )\n )\n\n # Save plot to BytesIO object\n img_bytes = BytesIO()\n fig.write_image(img_bytes, format=\"png\", scale=2)\n img_bytes.seek(0)\n \n # Return plot as PIL Image object\n return Image.open(img_bytes)"}}},{"rowIdx":3,"cells":{"id":{"kind":"string","value":"MatplotlibChartPipeline_gauge_2-30"},"image":{"kind":"image","value":{"src":"https://datasets-server.huggingface.co/assets/allenai/CoSyn-400K/--/86e46e1fd5e754d056169f0fb38f06c6997ff7de/--/chart/train/3/image/image.jpg?Expires=1741922929&Signature=k7qnjOAijYrD3BAs4I-hyFOF4M-SO8oZa0VZwrfVvhSoT-ffg4cctJbqfcdJ4gfAvlo2CqGaENTTaNlp7xWVppSqJEu9RgYim0whoiUqs4JPdOtes8sRpd5ldT6p3Rw6Ly9PrLcJagbSFW932OzuG8Lg-N~i7J~psuIihEulaOqs2TgIEpvP-WvN6bU9h0dCdrad~YqUGgVf40SoDjHLSoysdftCGgtxQpX7K81MBNTVYsOG8oGOepEpQ-dCBLOfjsoK2KL2t1yShtQ7h55Ic7YQsqOgvuNvjVd2ce0kGXzrTJKzV-0RVfRXmDd3APKz3CGX3Wr4WszI8s9vQAAMjw__&Key-Pair-Id=K3EI6M078Z3AC3","height":1222,"width":2560},"inferredFromUrl":false},"qa_pairs":{"kind":"list like","value":{"question":["What's the title of the figure?","What metric has the highest completion percentage?","What is the range of values shown in each gauge chart?","Which metric has the lowest progress percentage?","Calculate the difference between the highest and lowest progress percentages.","Which two metrics have values greater than 70%?","What is the average progress percentage of all the metrics?","Compare the Renewable Energy Adoption percentage with the Waste Recycling Rate percentage. Which is higher?","What is the most common color theme used in the gauge charts?","Which metric is closest to 50% progress?"],"explanation":["The title is written prominently at the top of the figure in large, bold font. It states the overall purpose of the visual.","The figure displays percentage values prominently in the middle of each gauge chart. The highest completion percentage is identified by comparing these numbers.","Each gauge chart has a scale from the minimum value to the maximum value, which are indicated near the ends of the gauge. By observing these scales, one can identify the range.","By examining the percentage values displayed in each gauge chart, the metric with the smallest percentage can be identified.","Identify the highest and lowest percentage values from the gauge charts. Subtract the lowest value from the highest value to get the difference. Highest is 89% (Employee Sustainability Training), and lowest is 32% (Carbon Footprint Reduction), so the difference is 89 - 32 = 57.","Check each gauge chart for values greater than 70%. Two charts have values of 71% and 89%.","Add all the given percentage values and divide by the number of metrics (7 metrics). (32 + 45 + 68 + 53 + 71 + 78 + 89) / 7 = 62.29.","Look at the percentage values for both metrics and compare them directly. Renewable Energy Adoption is 45% and Waste Recycling Rate is 68%, so Waste Recycling Rate is higher.","Look at the colors used in the gauge charts and see if a pattern or predominant color theme can be identified, which in this case is a gradient from the Viridis colormap.","Identify the percentage closest to 50% by comparing all the values. Water Conservation Progress is at 53%, which is closest."],"answer":["Sustainability Goals Progress","Employee Sustainability Training Completion (%)","0 to 100","Carbon Footprint Reduction (%)","57","Green Building Certification (%) and Employee Sustainability Training Completion (%)","62.29","Waste Recycling Rate (%)","Gradient from Viridis colormap","Water Conservation Progress (%)"]},"string":"{\n \"question\": [\n \"What's the title of the figure?\",\n \"What metric has the highest completion percentage?\",\n \"What is the range of values shown in each gauge chart?\",\n \"Which metric has the lowest progress percentage?\",\n \"Calculate the difference between the highest and lowest progress percentages.\",\n \"Which two metrics have values greater than 70%?\",\n \"What is the average progress percentage of all the metrics?\",\n \"Compare the Renewable Energy Adoption percentage with the Waste Recycling Rate percentage. Which is higher?\",\n \"What is the most common color theme used in the gauge charts?\",\n \"Which metric is closest to 50% progress?\"\n ],\n \"explanation\": [\n \"The title is written prominently at the top of the figure in large, bold font. It states the overall purpose of the visual.\",\n \"The figure displays percentage values prominently in the middle of each gauge chart. The highest completion percentage is identified by comparing these numbers.\",\n \"Each gauge chart has a scale from the minimum value to the maximum value, which are indicated near the ends of the gauge. By observing these scales, one can identify the range.\",\n \"By examining the percentage values displayed in each gauge chart, the metric with the smallest percentage can be identified.\",\n \"Identify the highest and lowest percentage values from the gauge charts. Subtract the lowest value from the highest value to get the difference. Highest is 89% (Employee Sustainability Training), and lowest is 32% (Carbon Footprint Reduction), so the difference is 89 - 32 = 57.\",\n \"Check each gauge chart for values greater than 70%. Two charts have values of 71% and 89%.\",\n \"Add all the given percentage values and divide by the number of metrics (7 metrics). (32 + 45 + 68 + 53 + 71 + 78 + 89) / 7 = 62.29.\",\n \"Look at the percentage values for both metrics and compare them directly. Renewable Energy Adoption is 45% and Waste Recycling Rate is 68%, so Waste Recycling Rate is higher.\",\n \"Look at the colors used in the gauge charts and see if a pattern or predominant color theme can be identified, which in this case is a gradient from the Viridis colormap.\",\n \"Identify the percentage closest to 50% by comparing all the values. Water Conservation Progress is at 53%, which is closest.\"\n ],\n \"answer\": [\n \"Sustainability Goals Progress\",\n \"Employee Sustainability Training Completion (%)\",\n \"0 to 100\",\n \"Carbon Footprint Reduction (%)\",\n \"57\",\n \"Green Building Certification (%) and Employee Sustainability Training Completion (%)\",\n \"62.29\",\n \"Waste Recycling Rate (%)\",\n \"Gradient from Viridis colormap\",\n \"Water Conservation Progress (%)\"\n ]\n}"},"metadata":{"kind":"string","value":"{\n \"figure_type\": \"Gauge Charts\",\n \"persona\": \"A prominent business leader who lends their expertise to enhance the university's curriculum\",\n \"topic\": \"Progress towards sustainability goals, including carbon footprint reduction and renewable energy adoption\"\n}"},"data":{"kind":"string","value":"Metric,Value,MinValue,MaxValue\nCarbon Footprint Reduction (%),32,0,100\nRenewable Energy Adoption (%),45,0,100\nWaste Recycling Rate (%),68,0,100\nWater Conservation Progress (%),53,0,100\nGreen Building Certification (%),71,0,100\nSustainable Procurement Score,78,0,100\nEmployee Sustainability Training Completion (%),89,0,100"},"code":{"kind":"string","value":"import numpy as np\nimport matplotlib.pyplot as plt\nfrom io import BytesIO\nfrom PIL import Image\n\ndef generate_plot(data):\n plt.style.use('Solarize_Light2')\n \n fig, axes = plt.subplots(2, 4, figsize=(20, 10), subplot_kw=dict(projection='polar'))\n fig.suptitle('Sustainability Goals Progress', fontsize=24, fontweight='bold', y=1.05)\n \n axes = axes.flatten()\n \n colors = plt.cm.viridis(np.linspace(0, 1, len(data)))\n \n for idx, (metric, value, min_value, max_value) in enumerate(data.itertuples(index=False)):\n ax = axes[idx]\n \n angle = np.radians(value / max_value * 180)\n r = 0.8\n \n ax.plot([0, angle], [0, r], color=colors[idx], linewidth=4)\n ax.set_ylim(0, 1)\n ax.set_yticks([])\n ax.set_xticks(np.radians([0, 45, 90, 135, 180]))\n ax.set_xticklabels([])\n \n ax.text(0, -0.2, f'{min_value}', ha='center', va='center', fontsize=10)\n ax.text(np.pi, -0.2, f'{max_value}', ha='center', va='center', fontsize=10)\n \n ax.text(0, 1.2, metric, ha='center', va='center', fontsize=12, fontweight='bold', wrap=True)\n ax.text(0, 0.5, f'{value}%', ha='center', va='center', fontsize=20, fontweight='bold')\n \n for spine in ax.spines.values():\n spine.set_visible(False)\n \n axes[-1].remove()\n \n plt.tight_layout()\n \n buffer = BytesIO()\n plt.savefig(buffer, format='png', dpi=300, bbox_inches='tight')\n buffer.seek(0)\n image = Image.open(buffer)\n \n return image"}}},{"rowIdx":4,"cells":{"id":{"kind":"string","value":"PlotlyChartPipeline_gauge_2-13"},"image":{"kind":"image","value":{"src":"https://datasets-server.huggingface.co/assets/allenai/CoSyn-400K/--/86e46e1fd5e754d056169f0fb38f06c6997ff7de/--/chart/train/4/image/image.jpg?Expires=1741922929&Signature=ZQPi8mzQDO9UxkeDdBloXizBcAO4MhW8haEG3NEXssPnsV4MtBGk2XdFeZG1JcdaeVOw5FLSrdQFmTNcK9RORGsIh8J38VoE476jccIqX3TCNy5lkRV5yklgzCBBbkLsv~AnRGnBRMPnsrw8zo86~vQNM2J8D27vWODIa~yKT0rBJSSTzHaTuAzWfJixD5Fbs-vM3877DZSk0b7CftA5defDZq00bjVzlIxwVJLs3k1fyx985~p~49DcD0TaAHfE4fvfW~iPAboQvE5eOEs~SkL4~won9QwkNbOn6GYE8qdjsWo1rdcU85yUJpUQmfXGg9fS~rfxx9rSKAHJ15pQSQ__&Key-Pair-Id=K3EI6M078Z3AC3","height":1440,"width":2160},"inferredFromUrl":false},"qa_pairs":{"kind":"list like","value":{"question":["What's the title of the chart?","Which ritual has the highest percentage of regular participation?","What is the percentage of the population that participates in weekly Shinto rituals?","How many categories are represented in the chart?","Compare the percentage of the population that visits Buddhist temples monthly to those who pray daily at home shrines. Which is higher and by how much?","What is the average percentage of the population participating in all the rituals shown?","Which two categories have the closest percentages?","What is the combined percentage of participants in annual and seasonal activities?","Are any categories below 20% participation? If yes, which ones?"],"explanation":["The title of the chart is found at the top of the figure. It reads \"日本の神社・寺院訪問習慣,\" which translates to \"Visitation Habits to Shrines and Temples in Japan.\"","By examining the gauge chart, the highest value displayed is 79.2%, which corresponds to \"Seasonal festival attendance.\"","Looking at the gauge chart, the gauge representing \"Weekly Shinto ritual participation\" shows a value of 15.7%.","The chart is divided into individual gauges per category. Counting the titles of each gauge, there are five categories.","\"Monthly visit to Buddhist temple\" shows 42.3%, and \"Daily home shrine prayer\" shows 31.8%. The difference is 42.3% - 31.8% = 10.5%.","Sum all percentages: 68.5 + 42.3 + 15.7 + 79.2 + 31.8 = 237.5. There are 5 categories, so the average is 237.5 / 5 = 47.5%.","Comparing all percentages: 68.5%, 42.3%, 15.7%, 79.2%, and 31.8%. The closest are \"Monthly visit to Buddhist temple\" at 42.3% and \"Daily home shrine prayer\" at 31.8%, with a difference of 10.5%.","Annual visit to local shrine is 68.5% and Seasonal festival attendance is 79.2%. Total combined percentage is 68.5% + 79.2% = 147.7%.","From the chart, the category \"Weekly Shinto ritual participation\" has a percentage of 15.7%, which is below 20%."],"answer":["日本の神社・寺院訪問習慣","Seasonal festival attendance","15.7%","5","Monthly visit to Buddhist temple is higher by 10.5%","47.5%","Monthly visit to Buddhist temple and Daily home shrine prayer","147.7%","Weekly Shinto ritual participation"]},"string":"{\n \"question\": [\n \"What's the title of the chart?\",\n \"Which ritual has the highest percentage of regular participation?\",\n \"What is the percentage of the population that participates in weekly Shinto rituals?\",\n \"How many categories are represented in the chart?\",\n \"Compare the percentage of the population that visits Buddhist temples monthly to those who pray daily at home shrines. Which is higher and by how much?\",\n \"What is the average percentage of the population participating in all the rituals shown?\",\n \"Which two categories have the closest percentages?\",\n \"What is the combined percentage of participants in annual and seasonal activities?\",\n \"Are any categories below 20% participation? If yes, which ones?\"\n ],\n \"explanation\": [\n \"The title of the chart is found at the top of the figure. It reads \\\"日本の神社・寺院訪問習慣,\\\" which translates to \\\"Visitation Habits to Shrines and Temples in Japan.\\\"\",\n \"By examining the gauge chart, the highest value displayed is 79.2%, which corresponds to \\\"Seasonal festival attendance.\\\"\",\n \"Looking at the gauge chart, the gauge representing \\\"Weekly Shinto ritual participation\\\" shows a value of 15.7%.\",\n \"The chart is divided into individual gauges per category. Counting the titles of each gauge, there are five categories.\",\n \"\\\"Monthly visit to Buddhist temple\\\" shows 42.3%, and \\\"Daily home shrine prayer\\\" shows 31.8%. The difference is 42.3% - 31.8% = 10.5%.\",\n \"Sum all percentages: 68.5 + 42.3 + 15.7 + 79.2 + 31.8 = 237.5. There are 5 categories, so the average is 237.5 / 5 = 47.5%.\",\n \"Comparing all percentages: 68.5%, 42.3%, 15.7%, 79.2%, and 31.8%. The closest are \\\"Monthly visit to Buddhist temple\\\" at 42.3% and \\\"Daily home shrine prayer\\\" at 31.8%, with a difference of 10.5%.\",\n \"Annual visit to local shrine is 68.5% and Seasonal festival attendance is 79.2%. Total combined percentage is 68.5% + 79.2% = 147.7%.\",\n \"From the chart, the category \\\"Weekly Shinto ritual participation\\\" has a percentage of 15.7%, which is below 20%.\"\n ],\n \"answer\": [\n \"日本の神社・寺院訪問習慣\",\n \"Seasonal festival attendance\",\n \"15.7%\",\n \"5\",\n \"Monthly visit to Buddhist temple is higher by 10.5%\",\n \"47.5%\",\n \"Monthly visit to Buddhist temple and Daily home shrine prayer\",\n \"147.7%\",\n \"Weekly Shinto ritual participation\"\n ]\n}"},"metadata":{"kind":"string","value":"{\n \"figure_type\": \"Gauge Chart (only display one number on each gauge without text overlapping)\",\n \"persona\": \"A fellow native Japanese speaker who shares her love for their homeland and often reminisces about their shared memories\",\n \"topic\": \"Percentage of Japanese population who regularly visit shrines or temples\"\n}"},"data":{"kind":"string","value":"Category,Percentage\nAnnual visit to local shrine,68.5\nMonthly visit to Buddhist temple,42.3\nWeekly Shinto ritual participation,15.7\nSeasonal festival attendance,79.2\nDaily home shrine prayer,31.8"},"code":{"kind":"string","value":"import plotly.graph_objects as go\nfrom plotly.subplots import make_subplots\nimport pandas as pd\nfrom io import BytesIO\nfrom PIL import Image\n\ndef generate_plot(data):\n fig = make_subplots(\n rows=2, cols=3,\n specs=[[{'type': 'indicator'}, {'type': 'indicator'}, {'type': 'indicator'}],\n [{'type': 'indicator'}, {'type': 'indicator'}, None]],\n subplot_titles=data['Category'].tolist(),\n )\n\n colors = ['#FF9999', '#66B2FF', '#99FF99', '#FFCC99', '#FF99CC']\n\n for i, (category, percentage) in enumerate(zip(data['Category'], data['Percentage'])):\n row = i // 3 + 1\n col = i % 3 + 1\n \n fig.add_trace(go.Indicator(\n mode=\"gauge+number\",\n value=percentage,\n domain={'row': row, 'column': col},\n title={'text': category, 'font': {'size': 16, 'family': 'Arial', 'color': colors[i]}},\n gauge={\n 'axis': {'range': [None, 100], 'tickwidth': 1, 'tickcolor': colors[i]},\n 'bar': {'color': colors[i]},\n 'bgcolor': \"white\",\n 'borderwidth': 2,\n 'bordercolor': colors[i],\n 'steps': [\n {'range': [0, percentage], 'color': colors[i]}\n ],\n },\n number={'font': {'size': 40, 'family': 'Arial', 'color': colors[i]}},\n ), row=row, col=col)\n\n fig.update_layout(\n title_text=\"日本の神社・寺院訪問習慣\",\n title_font={'size': 24, 'family': 'Arial', 'color': '#333333'},\n font_family=\"Arial\",\n font_color=\"#333333\",\n paper_bgcolor=\"#F9F9F9\",\n plot_bgcolor=\"#F9F9F9\",\n height=800,\n width=1200,\n margin=dict(l=50, r=50, t=100, b=50),\n )\n\n img_bytes = BytesIO()\n fig.write_image(img_bytes, format=\"png\", scale=2)\n img_bytes.seek(0)\n return Image.open(img_bytes)"}}},{"rowIdx":5,"cells":{"id":{"kind":"string","value":"MatplotlibChartPipeline_hexbin_3-377"},"image":{"kind":"image","value":{"src":"https://datasets-server.huggingface.co/assets/allenai/CoSyn-400K/--/86e46e1fd5e754d056169f0fb38f06c6997ff7de/--/chart/train/5/image/image.jpg?Expires=1741922929&Signature=hHreJDJrziTjimoGHqsLjGfued3vFrH1tK6cecDX9HpjPvjZNdgms4UFqXCP-1bFvjC~7pMlImHsF2dO4SfZggaNbRDH3RzZ1Zkj87rEy3JlhjO8HTNlTOWRc1UGDTzOgiMzviCVrhKHGK6z-AH2YJTpkttWMtd6OlfjZKgrRosSFGkLvr5S~u8Gpql4~4ztv1QLR3NDZuUkDpMXtpAOjeSBDMTalGzECx7obMRa~w0IB3Uy~nmsCBUrIOSpJGRjyHvFwflpGp~xT-4O8OCOPinBCLhMajoEydF3xjRyZK-DLJTnhlEm1wU1PBj7HyfS35wOHAy3huGuxcZeyLbyVA__&Key-Pair-Id=K3EI6M078Z3AC3","height":1440,"width":2021},"inferredFromUrl":false},"qa_pairs":{"kind":"list like","value":{"question":["What is the title of the figure?","What is the range of wait times shown on the x-axis?","Which department has the highest wait time for the 'Low' urgency level?","How many departments have a wait time of at least 180 minutes for the 'Low' urgency level?","What is the color representing the highest density of data points?","Between Pediatrics and Neurology, which department has a lower average wait time for the 'High' urgency level?","Which urgency level seems to exhibit the widest range of wait times across all departments?","How does the frequency distribution of 'Medium' urgency levels compare between Oncology and Radiology?","Which department has the closest frequency for 'High' urgency levels to the Emergency department?"],"explanation":["The title is prominently displayed at the top of the plot, designed to clearly state the main focus of the visualization.","By looking at the x-axis labels, you can see it spans from the minimum to maximum displayed wait times.","By following the labels on the plot, the department with the highest wait time for 'Low' urgency level can be identified.","Locate the 'Low' urgency level points in the plot and count the departments with wait times of 180 minutes or more.","The color bar adjacent to the plot shows the corresponding colors for different densities, where the color representing the highest density can be seen.","Identify the 'High' urgency level wait times for both departments, then calculate and compare their averages.","By comparing the spread of wait times (the range) for each urgency level, determine which one spans the widest interval.","Locate the wait times and frequencies for 'Medium' urgency in both Oncology and Radiology to compare their distributions.","Compare the frequencies of 'High' urgency levels for each department against the data for the Emergency department."],"answer":["Patient Wait Times vs Frequency","15 to 270 minutes","Orthopedics","Three departments","Dark purple","Pediatrics","Low","Radiology has a higher frequency for 'Medium' urgency","Radiology"]},"string":"{\n \"question\": [\n \"What is the title of the figure?\",\n \"What is the range of wait times shown on the x-axis?\",\n \"Which department has the highest wait time for the 'Low' urgency level?\",\n \"How many departments have a wait time of at least 180 minutes for the 'Low' urgency level?\",\n \"What is the color representing the highest density of data points?\",\n \"Between Pediatrics and Neurology, which department has a lower average wait time for the 'High' urgency level?\",\n \"Which urgency level seems to exhibit the widest range of wait times across all departments?\",\n \"How does the frequency distribution of 'Medium' urgency levels compare between Oncology and Radiology?\",\n \"Which department has the closest frequency for 'High' urgency levels to the Emergency department?\"\n ],\n \"explanation\": [\n \"The title is prominently displayed at the top of the plot, designed to clearly state the main focus of the visualization.\",\n \"By looking at the x-axis labels, you can see it spans from the minimum to maximum displayed wait times.\",\n \"By following the labels on the plot, the department with the highest wait time for 'Low' urgency level can be identified.\",\n \"Locate the 'Low' urgency level points in the plot and count the departments with wait times of 180 minutes or more.\",\n \"The color bar adjacent to the plot shows the corresponding colors for different densities, where the color representing the highest density can be seen.\",\n \"Identify the 'High' urgency level wait times for both departments, then calculate and compare their averages.\",\n \"By comparing the spread of wait times (the range) for each urgency level, determine which one spans the widest interval.\",\n \"Locate the wait times and frequencies for 'Medium' urgency in both Oncology and Radiology to compare their distributions.\",\n \"Compare the frequencies of 'High' urgency levels for each department against the data for the Emergency department.\"\n ],\n \"answer\": [\n \"Patient Wait Times vs Frequency\",\n \"15 to 270 minutes\",\n \"Orthopedics\",\n \"Three departments\",\n \"Dark purple\",\n \"Pediatrics\",\n \"Low\",\n \"Radiology has a higher frequency for 'Medium' urgency\",\n \"Radiology\"\n ]\n}"},"metadata":{"kind":"string","value":"{\n \"figure_type\": \"Hexbin Plot\",\n \"persona\": \"A nurse navigator who helps patients navigate the complex healthcare system and access necessary treatments\",\n \"topic\": \"Patient wait times across different departments in a hospital, broken down by urgency levels\"\n}"},"data":{"kind":"string","value":"Department,Urgency,Wait_Time_Minutes,Frequency\nEmergency,Critical,15,30\nEmergency,High,45,50\nEmergency,Medium,90,70\nEmergency,Low,180,40\nCardiology,High,60,25\nCardiology,Medium,120,45\nCardiology,Low,240,30\nOncology,High,45,20\nOncology,Medium,90,35\nOncology,Low,180,25\nOrthopedics,High,75,15\nOrthopedics,Medium,150,30\nOrthopedics,Low,270,20\nPediatrics,High,30,35\nPediatrics,Medium,75,55\nPediatrics,Low,150,40\nRadiology,High,40,30\nRadiology,Medium,90,50\nRadiology,Low,180,35\nNeurology,High,55,20\nNeurology,Medium,110,40\nNeurology,Low,220,25"},"code":{"kind":"string","value":"import numpy as np\nimport matplotlib.pyplot as plt\nimport io\nfrom PIL import Image\n\ndef generate_plot(data):\n plt.style.use('seaborn-v0_8-whitegrid')\n \n # Prepare data\n x = data['Wait_Time_Minutes']\n y = data['Frequency']\n \n # Create figure and axis\n fig, ax = plt.subplots(figsize=(12, 8))\n \n # Create hexbin plot\n hb = ax.hexbin(x, y, gridsize=15, cmap='viridis', edgecolors='white')\n \n # Customize plot\n ax.set_title('Patient Wait Times vs Frequency', fontsize=20, fontweight='bold', pad=20)\n ax.set_xlabel('Wait Time (Minutes)', fontsize=14, labelpad=15)\n ax.set_ylabel('Frequency', fontsize=14, labelpad=15)\n \n # Add colorbar\n cbar = plt.colorbar(hb, ax=ax)\n cbar.set_label('Count', rotation=270, labelpad=20, fontsize=12)\n \n # Customize ticks\n ax.tick_params(axis='both', which='major', labelsize=12)\n \n # Add text annotations for departments\n for _, row in data.iterrows():\n ax.annotate(row['Department'], (row['Wait_Time_Minutes'], row['Frequency']),\n xytext=(5, 5), textcoords='offset points', fontsize=8,\n bbox=dict(boxstyle=\"round,pad=0.3\", fc=\"white\", ec=\"gray\", alpha=0.8))\n \n # Adjust layout\n plt.tight_layout()\n \n # Save plot to BytesIO object\n buf = io.BytesIO()\n plt.savefig(buf, format='png', dpi=300, bbox_inches='tight')\n buf.seek(0)\n \n # Convert to PIL Image\n img = Image.open(buf)\n \n return img"}}},{"rowIdx":6,"cells":{"id":{"kind":"string","value":"PlotlyChartPipeline_line_claude_2-885"},"image":{"kind":"image","value":{"src":"https://datasets-server.huggingface.co/assets/allenai/CoSyn-400K/--/86e46e1fd5e754d056169f0fb38f06c6997ff7de/--/chart/train/6/image/image.jpg?Expires=1741922929&Signature=HlfR96ueo97mm2TeXbdJJwqQIyZNx1ZmKGRGQPvAxohEpZFD7E-kFOfrRPT3l1hlKuMWKchwuA1hPT8Xg~njCYqG5A77tpFdF29aUj7G57us3F5DznPe0k2xjnaZfvQmQBYFFxXQGkYs7aiq0t4wk3q9rAgkR0WnJVM2rCPCCTaLM-CtEHVw9ySJeGKDRsLvOcLBY-px6xWdvdmSXudaVj6vo3ik6la5qg4b1rvqDGc1jnIVQC23CKis7pWeJB0CS1KH1xF0HCzGaqMWrZzAId0zyZ-5CnkkjLeZnxvy8Jdncbx1JV4JtW5QMGxzV2md~MkdQfFz0mqzcL3jVo-IZA__&Key-Pair-Id=K3EI6M078Z3AC3","height":1440,"width":2160},"inferredFromUrl":false},"qa_pairs":{"kind":"list like","value":{"question":["What is the highest recorded price for Basil from 2018 to 2023?","What was the price trend for Bell Peppers from Jan 2018 to Jan 2023?","How do the prices of Cucumbers in Jan 2018 and Jan 2023 compare?","What is the average price of Tomatoes in 2019?","Which vegetable had the most consistent prices over the 5-year period?","Which vegetable had the steepest price increase between Jan 2020 and Jan 2023?","What can you say about the price pattern for Basil in 2021?","How does the price of Lettuce in Jul 2022 compare to the price in Jul 2018?"],"explanation":["To find the highest price for Basil, look at the Basil line in the plot and identify the peak point. From the chart, the highest point appears to be at the most recent date.","To determine the trend, observe the line representing Bell Peppers. Notice the fluctuations and whether the general direction is upward, downward, or stable. From the chart, the price of Bell Peppers generally increases over time.","Locate the points for Cucumbers in Jan 2018 and Jan 2023 on the plot. Compare their Y-axis values. In Jan 2018, the price is $2.79, and in Jan 2023, it is $3.79.","Identify the price points for Tomatoes in 2019 on the plot. These points are in Jan, Apr, Jul, and Oct. The prices are $3.69, $3.99, $3.49, and $3.79 respectively. Compute the average as (3.69 + 3.99 + 3.49 + 3.79) / 4 = 3.74.","Consistency in prices would be indicated by the least fluctuation in the respective line for each vegetable. Observing the chart, Lettuce seems to show the least variation in price over the 5-year period.","Look for the rate of change between Jan 2020 and Jan 2023 for each vegetable. Observe the differences in the starting and ending points for these dates for each line. Bell Peppers show the most significant increase from $4.69 to $5.59.","Observe the Basil line within the year 2021. Identify how the prices change from Jan, Apr, Jul, and Oct. The price starts at $3.59 in Jan, slightly decreases to $3.49 in Apr, then rises to $3.69 by Oct.","Locate the points for Lettuce in Jul 2022 and Jul 2018. Compare their values. In Jul 2022, the price is $3.59, and in Jul 2018, it is $2.39."],"answer":["$3.99","Increasing","Jan 2023 is higher","$3.74","Lettuce","Bell Peppers","Slightly fluctuating, then increasing","Jul 2022 is higher"]},"string":"{\n \"question\": [\n \"What is the highest recorded price for Basil from 2018 to 2023?\",\n \"What was the price trend for Bell Peppers from Jan 2018 to Jan 2023?\",\n \"How do the prices of Cucumbers in Jan 2018 and Jan 2023 compare?\",\n \"What is the average price of Tomatoes in 2019?\",\n \"Which vegetable had the most consistent prices over the 5-year period?\",\n \"Which vegetable had the steepest price increase between Jan 2020 and Jan 2023?\",\n \"What can you say about the price pattern for Basil in 2021?\",\n \"How does the price of Lettuce in Jul 2022 compare to the price in Jul 2018?\"\n ],\n \"explanation\": [\n \"To find the highest price for Basil, look at the Basil line in the plot and identify the peak point. From the chart, the highest point appears to be at the most recent date.\",\n \"To determine the trend, observe the line representing Bell Peppers. Notice the fluctuations and whether the general direction is upward, downward, or stable. From the chart, the price of Bell Peppers generally increases over time.\",\n \"Locate the points for Cucumbers in Jan 2018 and Jan 2023 on the plot. Compare their Y-axis values. In Jan 2018, the price is $2.79, and in Jan 2023, it is $3.79.\",\n \"Identify the price points for Tomatoes in 2019 on the plot. These points are in Jan, Apr, Jul, and Oct. The prices are $3.69, $3.99, $3.49, and $3.79 respectively. Compute the average as (3.69 + 3.99 + 3.49 + 3.79) / 4 = 3.74.\",\n \"Consistency in prices would be indicated by the least fluctuation in the respective line for each vegetable. Observing the chart, Lettuce seems to show the least variation in price over the 5-year period.\",\n \"Look for the rate of change between Jan 2020 and Jan 2023 for each vegetable. Observe the differences in the starting and ending points for these dates for each line. Bell Peppers show the most significant increase from $4.69 to $5.59.\",\n \"Observe the Basil line within the year 2021. Identify how the prices change from Jan, Apr, Jul, and Oct. The price starts at $3.59 in Jan, slightly decreases to $3.49 in Apr, then rises to $3.69 by Oct.\",\n \"Locate the points for Lettuce in Jul 2022 and Jul 2018. Compare their values. In Jul 2022, the price is $3.59, and in Jul 2018, it is $2.39.\"\n ],\n \"answer\": [\n \"$3.99\",\n \"Increasing\",\n \"Jan 2023 is higher\",\n \"$3.74\",\n \"Lettuce\",\n \"Bell Peppers\",\n \"Slightly fluctuating, then increasing\",\n \"Jul 2022 is higher\"\n ]\n}"},"metadata":{"kind":"string","value":"{\n \"figure_type\": \"line chart\",\n \"persona\": \"An urban restaurant owner with a dream of having a rooftop garden to provide fresh herbs and vegetables for his menu\",\n \"topic\": \"Monthly average prices of organic vegetables in local markets over 5 years\"\n}"},"data":{"kind":"string","value":"Month,Basil,Tomatoes,Lettuce,Bell Peppers,Cucumbers\nJan 2018,2.99,3.49,2.29,3.99,2.79\nApr 2018,2.89,3.79,2.19,4.29,2.69\nJul 2018,2.79,3.29,2.39,3.89,2.59\nOct 2018,3.09,3.59,2.49,4.19,2.89\nJan 2019,3.19,3.69,2.59,4.39,2.99\nApr 2019,3.09,3.99,2.49,4.59,2.89\nJul 2019,2.99,3.49,2.69,4.09,2.79\nOct 2019,3.29,3.79,2.79,4.49,3.09\nJan 2020,3.39,3.89,2.89,4.69,3.19\nApr 2020,3.29,4.19,2.79,4.89,3.09\nJul 2020,3.19,3.69,2.99,4.39,2.99\nOct 2020,3.49,3.99,3.09,4.79,3.29\nJan 2021,3.59,4.09,3.19,4.99,3.39\nApr 2021,3.49,4.39,3.09,5.19,3.29\nJul 2021,3.39,3.89,3.29,4.69,3.19\nOct 2021,3.69,4.19,3.39,5.09,3.49\nJan 2022,3.79,4.29,3.49,5.29,3.59\nApr 2022,3.69,4.59,3.39,5.49,3.49\nJul 2022,3.59,4.09,3.59,4.99,3.39\nOct 2022,3.89,4.39,3.69,5.39,3.69\nJan 2023,3.99,4.49,3.79,5.59,3.79"},"code":{"kind":"string","value":"import plotly.graph_objects as go\nimport pandas as pd\nfrom io import BytesIO\nfrom PIL import Image\n\ndef generate_plot(df):\n # Convert 'Month' column to datetime\n df['Month'] = pd.to_datetime(df['Month'], format='%b %Y')\n\n # Create the figure\n fig = go.Figure()\n\n # Add traces for each vegetable\n colors = ['#FF6B6B', '#4ECDC4', '#45B7D1', '#FFA07A', '#98D8C8']\n for i, col in enumerate(df.columns[1:]):\n fig.add_trace(go.Scatter(\n x=df['Month'],\n y=df[col],\n name=col,\n line=dict(color=colors[i], width=3),\n mode='lines+markers',\n marker=dict(size=8, symbol='diamond', line=dict(width=2, color='white'))\n ))\n\n # Update layout\n fig.update_layout(\n title={\n 'text': 'Monthly Average Prices of Organic Vegetables (2018-2023)',\n 'y':0.95,\n 'x':0.5,\n 'xanchor': 'center',\n 'yanchor': 'top',\n 'font': dict(size=24, family='Arial', color='#333333')\n },\n xaxis_title='Date',\n yaxis_title='Price (USD)',\n legend_title='Vegetables',\n plot_bgcolor='rgba(240,240,240,0.95)',\n font=dict(family='Arial', size=14, color='#333333'),\n xaxis=dict(\n showgrid=True,\n gridcolor='rgba(200,200,200,0.2)',\n tickangle=45,\n tickformat='%b %Y'\n ),\n yaxis=dict(\n showgrid=True,\n gridcolor='rgba(200,200,200,0.2)',\n tickprefix='$'\n ),\n legend=dict(\n bgcolor='rgba(255,255,255,0.5)',\n bordercolor='rgba(200,200,200,0.7)',\n borderwidth=1\n ),\n hovermode='x unified',\n margin=dict(l=80, r=80, t=100, b=80)\n )\n\n # Add a rectangle shape to highlight the most recent data point\n last_date = df['Month'].iloc[-1]\n fig.add_shape(\n type=\"rect\",\n xref=\"x\", yref=\"paper\",\n x0=last_date, y0=0, x1=last_date, y1=1,\n fillcolor=\"rgba(255,255,0,0.3)\",\n layer=\"below\",\n line_width=0,\n )\n\n # Save the plot to BytesIO\n img_bytes = BytesIO()\n fig.write_image(img_bytes, format='png', width=1200, height=800, scale=2)\n img_bytes.seek(0)\n \n # Convert to PIL Image\n img = Image.open(img_bytes)\n \n return img"}}},{"rowIdx":7,"cells":{"id":{"kind":"string","value":"MatplotlibChartPipeline_hexbin_2-66"},"image":{"kind":"image","value":{"src":"https://datasets-server.huggingface.co/assets/allenai/CoSyn-400K/--/86e46e1fd5e754d056169f0fb38f06c6997ff7de/--/chart/train/7/image/image.jpg?Expires=1741922929&Signature=NSRtHWcKG6NLfhtr~Mzb5tzJIezPWP-Ntiv2TeeP1OBtwXi3p7gpFRZpSWz4M6FIOn7t2TJeWCsIcppHcNmNDad8vRXhCpF4uR3UB-J5fGQtegTd-s3yNOT04IUft43J8q3ACVl~5FtfgSg-0x1lnUtJ8tKbPDk8keQjILOhurYJsGkskVVNBqFUYm3pqnKLmIqFnvGQNzBGUpTS8x7W9EBHqEwHe9UdybvH6Zj4Ux5dPdaTI-lJn7E2BJqfsf2~26K7QfkLf3wR9Xj46FUZ~EdIWKVEZCmdsYTOHwKsDPpN6XnUcpdtEQmZ6kWA2S1Vp3-8HgY9inX8wY6M9TFCAA__&Key-Pair-Id=K3EI6M078Z3AC3","height":1440,"width":1702},"inferredFromUrl":false},"qa_pairs":{"kind":"list like","value":{"question":["What is the title of the figure?","What are the labels of the x-axis and y-axis?","Which part of the color spectrum in the hexbin plot represents the highest frequency of quantum superposition states?","What is the frequency range shown in the color bar?","How many hexagons are plotted in the hexbin plot?","Which coordinate region has the highest density of quantum superposition states?","Is there a higher frequency of quantum superposition states at lower x-coordinates (0.2 to 0.5) compared to higher x-coordinates (0.8 to 1.2)?","Between x=0.4, y=0.5 and x=1.2, y=1.3, which coordinate has a higher frequency?","Is the distribution of the frequencies in the hexbin plot more concentrated or spread out?","What can be inferred about the relationship between x and y coordinates in terms of frequency?"],"explanation":["The title is located at the top-center of the figure and is usually styled to catch attention, it reads: \"Probability Distribution of Quantum Superposition States in Imaginary Alien Technologies\"","Typically, axis labels are found along the horizontal and vertical axes. In this figure, the x-axis label is 'X Coordinate' and the y-axis label is 'Y Coordinate'.","The color spectrum in the hexbin plot ranges from dark to light. The lightest color (yellowish) indicates the highest frequency, as seen in some of the central hexagons.","The frequency range is displayed on the color bar, which lies beside the plot. It starts at the lowest value shown and ends at the highest value.","Count the visible hexagons within the plotted area on the figure. Each hexagon represents a bin of data points.","The highest density region can be identified by locating the hexagon with the brightest color, indicating the highest frequency. This region is around x=0.7, y=1.0","To answer, compare the number and color intensity of the hexagons in the mentioned x-coordinate ranges. Observe which range has more high-frequency hexagons.","Compare the frequency values represented by the color intensity of the hexagons at these coordinates.","Observe the pattern of the hexbin plot, determining whether the higher frequency areas are centralized or dispersed across the plot.","Analyze whether there is a trend in frequency changes along the x and y coordinates, such as whether higher x-coordinates correlate with higher y-coordinates and higher frequencies."],"answer":["Probability Distribution of Quantum Superposition States in Imaginary Alien Technologies","X Coordinate and Y Coordinate","Yellowish color","10 to 35","20","Around x=0.7 and y=1.0","Yes, higher frequency at higher x-coordinates (0.8 to 1.2)","x=0.4, y=0.5","Concentrated","Higher x-coordinates and y-coordinates tend to have higher frequencies"]},"string":"{\n \"question\": [\n \"What is the title of the figure?\",\n \"What are the labels of the x-axis and y-axis?\",\n \"Which part of the color spectrum in the hexbin plot represents the highest frequency of quantum superposition states?\",\n \"What is the frequency range shown in the color bar?\",\n \"How many hexagons are plotted in the hexbin plot?\",\n \"Which coordinate region has the highest density of quantum superposition states?\",\n \"Is there a higher frequency of quantum superposition states at lower x-coordinates (0.2 to 0.5) compared to higher x-coordinates (0.8 to 1.2)?\",\n \"Between x=0.4, y=0.5 and x=1.2, y=1.3, which coordinate has a higher frequency?\",\n \"Is the distribution of the frequencies in the hexbin plot more concentrated or spread out?\",\n \"What can be inferred about the relationship between x and y coordinates in terms of frequency?\"\n ],\n \"explanation\": [\n \"The title is located at the top-center of the figure and is usually styled to catch attention, it reads: \\\"Probability Distribution of Quantum Superposition States in Imaginary Alien Technologies\\\"\",\n \"Typically, axis labels are found along the horizontal and vertical axes. In this figure, the x-axis label is 'X Coordinate' and the y-axis label is 'Y Coordinate'.\",\n \"The color spectrum in the hexbin plot ranges from dark to light. The lightest color (yellowish) indicates the highest frequency, as seen in some of the central hexagons.\",\n \"The frequency range is displayed on the color bar, which lies beside the plot. It starts at the lowest value shown and ends at the highest value.\",\n \"Count the visible hexagons within the plotted area on the figure. Each hexagon represents a bin of data points.\",\n \"The highest density region can be identified by locating the hexagon with the brightest color, indicating the highest frequency. This region is around x=0.7, y=1.0\",\n \"To answer, compare the number and color intensity of the hexagons in the mentioned x-coordinate ranges. Observe which range has more high-frequency hexagons.\",\n \"Compare the frequency values represented by the color intensity of the hexagons at these coordinates.\",\n \"Observe the pattern of the hexbin plot, determining whether the higher frequency areas are centralized or dispersed across the plot.\",\n \"Analyze whether there is a trend in frequency changes along the x and y coordinates, such as whether higher x-coordinates correlate with higher y-coordinates and higher frequencies.\"\n ],\n \"answer\": [\n \"Probability Distribution of Quantum Superposition States in Imaginary Alien Technologies\",\n \"X Coordinate and Y Coordinate\",\n \"Yellowish color\",\n \"10 to 35\",\n \"20\",\n \"Around x=0.7 and y=1.0\",\n \"Yes, higher frequency at higher x-coordinates (0.8 to 1.2)\",\n \"x=0.4, y=0.5\",\n \"Concentrated\",\n \"Higher x-coordinates and y-coordinates tend to have higher frequencies\"\n ]\n}"},"metadata":{"kind":"string","value":"{\n \"figure_type\": \"Hexbin Plot\",\n \"persona\": \"A theoretical physicist who collaborated with the persona on research related to quantum theories in science fiction\",\n \"topic\": \"Probability distribution of quantum superposition states in imaginary alien technologies\"\n}"},"data":{"kind":"string","value":"x,y,frequency\n0.2,0.3,15\n0.4,0.5,22\n0.6,0.7,30\n0.8,0.9,18\n1.0,1.1,25\n1.2,1.3,12\n0.3,0.6,20\n0.5,0.8,28\n0.7,1.0,35\n0.9,1.2,16\n1.1,1.4,23\n0.1,0.4,10\n0.3,0.7,19\n0.5,1.0,27\n0.7,1.3,32\n0.9,1.5,14\n1.1,1.7,21\n0.2,0.8,17\n0.4,1.1,24\n0.6,1.4,29"},"code":{"kind":"string","value":"import numpy as np\nimport matplotlib.pyplot as plt\nfrom io import BytesIO\nfrom PIL import Image\n\ndef generate_plot(df):\n plt.style.use('_classic_test_patch')\n \n fig, ax = plt.subplots(figsize=(10, 8))\n \n hex_plot = ax.hexbin(df['x'], df['y'], C=df['frequency'], \n gridsize=10, cmap='viridis', \n edgecolors='white', linewidths=0.4)\n \n ax.set_title('Probability Distribution of Quantum Superposition States\\nin Imaginary Alien Technologies', \n fontsize=16, fontweight='bold', color='#444444')\n ax.set_xlabel('X Coordinate', fontsize=12, fontweight='bold', color='#555555')\n ax.set_ylabel('Y Coordinate', fontsize=12, fontweight='bold', color='#555555')\n \n ax.set_xlim(0, 1.3)\n ax.set_ylim(0.2, 1.8)\n \n ax.tick_params(axis='both', which='major', labelsize=10, colors='#666666')\n \n cbar = plt.colorbar(hex_plot, ax=ax)\n cbar.set_label('Frequency', fontsize=12, fontweight='bold', color='#555555')\n cbar.ax.tick_params(labelsize=10, colors='#666666')\n \n ax.grid(True, linestyle='--', alpha=0.6)\n ax.set_facecolor('#f0f0f0')\n \n plt.tight_layout()\n \n buffer = BytesIO()\n plt.savefig(buffer, format='png', dpi=300, bbox_inches='tight')\n buffer.seek(0)\n image = Image.open(buffer)\n \n return image"}}},{"rowIdx":8,"cells":{"id":{"kind":"string","value":"PlotlyChartPipeline_funnel_1-319"},"image":{"kind":"image","value":{"src":"https://datasets-server.huggingface.co/assets/allenai/CoSyn-400K/--/86e46e1fd5e754d056169f0fb38f06c6997ff7de/--/chart/train/8/image/image.jpg?Expires=1741922929&Signature=jUbPDJIM7DZSRAEJWAk~fZ3wYr3VXIXqyAo2FZ4t40yayqTQtot3S3vks999Sxox9YmaEw72iU4fcNdBRzRVNsir6EXOZPplJ0DFUsOrDwvk-2dABui3l9Gt~XNgtqx26vbxQgDSKBhaSd9i9txvrvDhYJCPLY9eZR1zi1-rXNPcQsMYByL~Oy4QrLtepsE9ETRWxeqDiOk5sONDvffOtoKmHuiXftNZA3UV03q6N89Jrc6QHHZO56092yOjCbqJQqsLi0mVUE~ZgHbPxjDgGIPRD2kS31YcvsKioy5mMRS55kZddhVauRbGiviOyqQHLCQxLff71DX2-wN6pL5Uqw__&Key-Pair-Id=K3EI6M078Z3AC3","height":1600,"width":2000},"inferredFromUrl":false},"qa_pairs":{"kind":"list like","value":{"question":["what is the title of the funnel chart?","How many phases are presented in the funnel chart?","Which phase sees the largest drop in percentage from the previous phase?","What percentage of people experience emotional support needs during the 'Networking and Community Building' phase?","What is the combined percentage of people who need emotional support in 'Career Transition Anxiety' and 'Celebrating Milestones'?","Between which two phases is the drop in percentage the smallest?","At what phase does the percentage of people needing emotional support fall below half of the initial value?","What is the color of the segment for the 'Dealing with Technical Challenges' phase?","How much percentage does the 'Balancing Work-Life-Study' phase lose compared to the 'Initial Learning Excitement' phase?","Which phase shows a greater support need: 'Networking and Community Building' or 'Career Transition Anxiety'?"],"explanation":["The title of the funnel chart is located at the top and is written in a large, dark blue font.","To determine the number of phases, count the distinct horizontal segments in the funnel chart.","Compare the percentage difference between consecutive phases to find the largest drop. The steepest drop occurs between Balancing Work-Life-Study (85%) and Overcoming Self-Doubt (70%).","Find the phase labeled 'Networking and Community Building' and refer to the percentage value associated with it.","Add the percentage values for the phases 'Career Transition Anxiety' and 'Celebrating Milestones': 30% + 20% = 50%.","Compare the percentage differences between each consecutive phase. The smallest drop is between Overcoming Self-Doubt (70%) and Dealing with Technical Challenges (60%), which is a 10% difference.","Find when the percentage value drops below 50% of the initial value (100%). This happens at the 'Maintaining Motivation' phase, which is 50%.","Identify the color associated with the 'Dealing with Technical Challenges' phase by looking at the funnel chart.","Subtract the percentage of 'Balancing Work-Life-Study' from 'Initial Learning Excitement': 100% - 85% = 15%.","Directly compare the percentages listed for 'Networking and Community Building' (40%) and 'Career Transition Anxiety' (30%)."],"answer":["Emotional Support Needs in Coding Journey for Single Parents","8","Balancing Work-Life-Study to Overcoming Self-Doubt","40%","50%","Overcoming Self-Doubt and Dealing with Technical Challenges","Maintaining Motivation","Purple (#DDA0DD)","15%","Networking and Community Building"]},"string":"{\n \"question\": [\n \"what is the title of the funnel chart?\",\n \"How many phases are presented in the funnel chart?\",\n \"Which phase sees the largest drop in percentage from the previous phase?\",\n \"What percentage of people experience emotional support needs during the 'Networking and Community Building' phase?\",\n \"What is the combined percentage of people who need emotional support in 'Career Transition Anxiety' and 'Celebrating Milestones'?\",\n \"Between which two phases is the drop in percentage the smallest?\",\n \"At what phase does the percentage of people needing emotional support fall below half of the initial value?\",\n \"What is the color of the segment for the 'Dealing with Technical Challenges' phase?\",\n \"How much percentage does the 'Balancing Work-Life-Study' phase lose compared to the 'Initial Learning Excitement' phase?\",\n \"Which phase shows a greater support need: 'Networking and Community Building' or 'Career Transition Anxiety'?\"\n ],\n \"explanation\": [\n \"The title of the funnel chart is located at the top and is written in a large, dark blue font.\",\n \"To determine the number of phases, count the distinct horizontal segments in the funnel chart.\",\n \"Compare the percentage difference between consecutive phases to find the largest drop. The steepest drop occurs between Balancing Work-Life-Study (85%) and Overcoming Self-Doubt (70%).\",\n \"Find the phase labeled 'Networking and Community Building' and refer to the percentage value associated with it.\",\n \"Add the percentage values for the phases 'Career Transition Anxiety' and 'Celebrating Milestones': 30% + 20% = 50%.\",\n \"Compare the percentage differences between each consecutive phase. The smallest drop is between Overcoming Self-Doubt (70%) and Dealing with Technical Challenges (60%), which is a 10% difference.\",\n \"Find when the percentage value drops below 50% of the initial value (100%). This happens at the 'Maintaining Motivation' phase, which is 50%.\",\n \"Identify the color associated with the 'Dealing with Technical Challenges' phase by looking at the funnel chart.\",\n \"Subtract the percentage of 'Balancing Work-Life-Study' from 'Initial Learning Excitement': 100% - 85% = 15%.\",\n \"Directly compare the percentages listed for 'Networking and Community Building' (40%) and 'Career Transition Anxiety' (30%).\"\n ],\n \"answer\": [\n \"Emotional Support Needs in Coding Journey for Single Parents\",\n \"8\",\n \"Balancing Work-Life-Study to Overcoming Self-Doubt\",\n \"40%\",\n \"50%\",\n \"Overcoming Self-Doubt and Dealing with Technical Challenges\",\n \"Maintaining Motivation\",\n \"Purple (#DDA0DD)\",\n \"15%\",\n \"Networking and Community Building\"\n ]\n}"},"metadata":{"kind":"string","value":"{\n \"figure_type\": \"Funnel Chart\",\n \"persona\": \"A single parent who understands the challenges and provides emotional support throughout the coding journey\",\n \"topic\": \"Emotional support needs at different phases of the coding journey for single parents\"\n}"},"data":{"kind":"string","value":"Phase,Percentage\nInitial Learning Excitement,100\nBalancing Work-Life-Study,85\nOvercoming Self-Doubt,70\nDealing with Technical Challenges,60\nMaintaining Motivation,50\nNetworking and Community Building,40\nCareer Transition Anxiety,30\nCelebrating Milestones,20"},"code":{"kind":"string","value":"import plotly.graph_objects as go\nimport plotly.io as pio\nfrom io import BytesIO\nfrom PIL import Image\nimport pandas as pd\n\ndef generate_plot(data):\n # Reverse the order of data for the funnel chart\n data = data.iloc[::-1].reset_index(drop=True)\n\n # Create the funnel chart\n fig = go.Figure(go.Funnel(\n y = data['Phase'],\n x = data['Percentage'],\n textposition = \"inside\",\n textinfo = \"value+percent initial\",\n opacity = 0.65,\n marker = {\"color\": [\"#FFA07A\", \"#98FB98\", \"#87CEFA\", \"#DDA0DD\", \"#F0E68C\", \"#00CED1\", \"#FF6347\", \"#7B68EE\"],\n \"line\": {\"width\": [4, 2, 2, 3, 1, 1, 1, 1], \"color\": [\"wheat\", \"wheat\", \"blue\", \"wheat\", \"wheat\", \"wheat\", \"orange\", \"wheat\"]}},\n connector = {\"line\": {\"color\": \"royalblue\", \"dash\": \"dot\", \"width\": 3}}\n ))\n\n # Update the layout\n fig.update_layout(\n title = {\n 'text': \"Emotional Support Needs in Coding Journey for Single Parents\",\n 'y':0.95,\n 'x':0.5,\n 'xanchor': 'center',\n 'yanchor': 'top',\n 'font': {'size': 24, 'color': 'darkblue', 'family': \"Arial, sans-serif\"}\n },\n font = {'family': \"Arial, sans-serif\", 'size': 14},\n paper_bgcolor = 'rgba(0,0,0,0)',\n plot_bgcolor = 'rgba(0,0,0,0)',\n height = 800,\n width = 1000,\n margin = dict(t=100, l=50, r=50, b=50)\n )\n\n # Save the plot to BytesIO\n img_bytes = BytesIO()\n pio.write_image(fig, img_bytes, format='png', scale=2)\n img_bytes.seek(0)\n \n # Return as PIL Image\n return Image.open(img_bytes)"}}},{"rowIdx":9,"cells":{"id":{"kind":"string","value":"MatplotlibChart_bar_claude_2-885"},"image":{"kind":"image","value":{"src":"https://datasets-server.huggingface.co/assets/allenai/CoSyn-400K/--/86e46e1fd5e754d056169f0fb38f06c6997ff7de/--/chart/train/9/image/image.jpg?Expires=1741922929&Signature=gWjQq~jNfCYQnVmgFFC6qnuM0Kou2CWKJEzbdWlvvDj6mRK-etSRTA7FFVL7QGIsClBIOW6e8IFH4LmhY7nzeqWfwiVBEUFUFhF8-pTTcjZ4NVgYEDRhUWup9KEHOC~7eg9ataDNgHmZhDpGdFxWKuQ3OJJU5CCWs8za1AxDXzWmR3iOFqwPSE83PLF1ih5OLJl5bNeeDqqJhjwaIiDCKf-eIRpFMAetDhBEGVDo4lhKzJ2d2o5vR2jcQyDJFexMPusfT1j0Wg0sFDDYXBCbiCeCkW6cUELh8ws0Zp~izE-NWc1XB-zXv3pWjv7EngqUUdkv8y0FkT24P5drGR6qPA__&Key-Pair-Id=K3EI6M078Z3AC3","height":1440,"width":2169},"inferredFromUrl":false},"qa_pairs":{"kind":"list like","value":{"question":["What's the percentage change in donors contributing under $100 from 2018 to 2022?","Which year had the highest percentage of donors contributing 100-$499?","Compare the percentage of donors contributing $1000-$4999 in 2018 and 2021. Which year had a higher percentage and by how much?","What was the combined percentage of donors contributing $500-$999 and $10000+ in 2019?","In which year did the $5000-$9999 contribution level see its highest percentage, and what was that percentage?","Which contribution level showed the most consistent percentage across the years 2018-2022?","What was the difference in percentage of donors contributing under $100 between 2020 and 2021?","What's the average percentage of donors contributing $500-$999 across the 5 years?","Which year had the lowest overall percentage of donors contributing $5000-$9999, and what was it?","Was there a year where the percentage of donors contributing under $100 was equal to those contributing $500-$999?"],"explanation":["To find the percentage change, subtract the percentage for 2022 from the percentage for 2018 and then divide by the percentage for 2018, finally multiply by 100. (32% - 26%) / 32% * 100 = 18.75%","From the bar chart, we can see that 2022 had the highest percentage of donors in the 100-$499 category, marked by the height of the green bar.","Looking at the heights of the orange bars, the percentage for 2018 is 14% and for 2021 it is also 14%. So, the difference is 0%.","Add the percentages for these categories in 2019: 19% (for $500-$999) + 2% (for $10000+). 19% + 2% = 21%.","By examining the purple bars, 2020 had the highest percentage for this level, marked at 7%.","By observing the bars and their heights, the $10000+ contribution level shows a consistent percentage of 2% across all years.","From the blue bars, the values are 35% (2020) and 28% (2021). The difference is 35% - 28% = 7%.","Add the percentages for these years and divide by 5: (18% + 19% + 17% + 20% + 21%) / 5. The sum is 95%, so the average is 95% / 5 = 19%.","By examining the purple bars, the lowest percentage is for 2019 and 2021, both marked at 5%.","No such year has equal percentages when comparing the blue bars for under $100 and the yellow bars for $500-$999."],"answer":["18.75%","2022","Both years are equal","21%","2020, 7%","$10000+","7%","19%","2019 and 2021, 5%","No"]},"string":"{\n \"question\": [\n \"What's the percentage change in donors contributing under $100 from 2018 to 2022?\",\n \"Which year had the highest percentage of donors contributing 100-$499?\",\n \"Compare the percentage of donors contributing $1000-$4999 in 2018 and 2021. Which year had a higher percentage and by how much?\",\n \"What was the combined percentage of donors contributing $500-$999 and $10000+ in 2019?\",\n \"In which year did the $5000-$9999 contribution level see its highest percentage, and what was that percentage?\",\n \"Which contribution level showed the most consistent percentage across the years 2018-2022?\",\n \"What was the difference in percentage of donors contributing under $100 between 2020 and 2021?\",\n \"What's the average percentage of donors contributing $500-$999 across the 5 years?\",\n \"Which year had the lowest overall percentage of donors contributing $5000-$9999, and what was it?\",\n \"Was there a year where the percentage of donors contributing under $100 was equal to those contributing $500-$999?\"\n ],\n \"explanation\": [\n \"To find the percentage change, subtract the percentage for 2022 from the percentage for 2018 and then divide by the percentage for 2018, finally multiply by 100. (32% - 26%) / 32% * 100 = 18.75%\",\n \"From the bar chart, we can see that 2022 had the highest percentage of donors in the 100-$499 category, marked by the height of the green bar.\",\n \"Looking at the heights of the orange bars, the percentage for 2018 is 14% and for 2021 it is also 14%. So, the difference is 0%.\",\n \"Add the percentages for these categories in 2019: 19% (for $500-$999) + 2% (for $10000+). 19% + 2% = 21%.\",\n \"By examining the purple bars, 2020 had the highest percentage for this level, marked at 7%.\",\n \"By observing the bars and their heights, the $10000+ contribution level shows a consistent percentage of 2% across all years.\",\n \"From the blue bars, the values are 35% (2020) and 28% (2021). The difference is 35% - 28% = 7%.\",\n \"Add the percentages for these years and divide by 5: (18% + 19% + 17% + 20% + 21%) / 5. The sum is 95%, so the average is 95% / 5 = 19%.\",\n \"By examining the purple bars, the lowest percentage is for 2019 and 2021, both marked at 5%.\",\n \"No such year has equal percentages when comparing the blue bars for under $100 and the yellow bars for $500-$999.\"\n ],\n \"answer\": [\n \"18.75%\",\n \"2022\",\n \"Both years are equal\",\n \"21%\",\n \"2020, 7%\",\n \"$10000+\",\n \"7%\",\n \"19%\",\n \"2019 and 2021, 5%\",\n \"No\"\n ]\n}"},"metadata":{"kind":"string","value":"{\n \"figure_type\": \"bar chart\",\n \"persona\": \"A nonprofit organization leader who commissions custom cakes for fundraising events\",\n \"topic\": \"Donor contribution levels for nonprofit fundraisers over the past 5 years\"\n}"},"data":{"kind":"string","value":"Year,Under $100,100-$499,$500-$999,$1000-$4999,$5000-$9999,$10000+\n2018,32%,28%,18%,14%,6%,2%\n2019,30%,29%,19%,15%,5%,2%\n2020,35%,26%,17%,13%,7%,2%\n2021,28%,31%,20%,14%,5%,2%\n2022,26%,33%,21%,13%,5%,2%"},"code":{"kind":"string","value":"import numpy as np\nimport matplotlib.pyplot as plt\nfrom io import BytesIO\nfrom PIL import Image\n\ndef generate_plot(data):\n plt.style.use('seaborn-v0_8-whitegrid')\n \n years = data['Year'].tolist()\n contribution_levels = data.columns[1:].tolist()\n \n fig, ax = plt.subplots(figsize=(12, 8))\n \n bar_width = 0.13\n opacity = 0.8\n index = np.arange(len(years))\n \n colors = ['#FF9999', '#66B2FF', '#99FF99', '#FFCC99', '#FF99CC', '#FFFF99']\n \n for i, level in enumerate(contribution_levels):\n values = [float(x.strip('%')) for x in data[level]]\n ax.bar(index + i*bar_width, values, bar_width,\n alpha=opacity, color=colors[i], label=level)\n \n ax.set_xlabel('Year', fontsize=14, fontweight='bold')\n ax.set_ylabel('Percentage of Donors', fontsize=14, fontweight='bold')\n ax.set_title('Donor Contribution Levels (2018-2022)', fontsize=18, fontweight='bold')\n ax.set_xticks(index + 2.5*bar_width)\n ax.set_xticklabels(years, fontsize=12)\n ax.set_ylim(0, 40)\n ax.legend(fontsize=10, loc='upper left', bbox_to_anchor=(1, 1))\n \n plt.setp(ax.get_yticklabels(), fontsize=12)\n \n for spine in ax.spines.values():\n spine.set_visible(False)\n \n ax.yaxis.grid(True, linestyle='--', alpha=0.7)\n \n plt.tight_layout()\n \n buffer = BytesIO()\n plt.savefig(buffer, format='png', dpi=300, bbox_inches='tight')\n buffer.seek(0)\n image = Image.open(buffer)\n \n return image"}}},{"rowIdx":10,"cells":{"id":{"kind":"string","value":"MatplotlibChartPipeline_bubble_1-278"},"image":{"kind":"image","value":{"src":"https://datasets-server.huggingface.co/assets/allenai/CoSyn-400K/--/86e46e1fd5e754d056169f0fb38f06c6997ff7de/--/chart/train/10/image/image.jpg?Expires=1741922929&Signature=r4Q5~XqE2wZthxb10QTWIaOHQlde6SG1B9En41HNHVGwWnwWrF-p0IsgHxkvVBgWSr0M3JB1vTwvoEAy3Ih05X5bRVs1vlrIvurGfUB~VRZVXFhlo2bZOB66NC3LdKlJpnc0a4uNQofLmmLrVivdQS0CaK~afdV1QDB5zBU2u15~RIE6eoU6HuAWjvSIzPlt8n0YpwsE~E8YEbQ2CiLbAE01oHCpn3~LDN~SQsflOVJ8AInUrkiky4PFAIh7XRJ5021QVim8Y~k00rHO4EbeNfU4Ri1j-ROcwOjiVEaBeVVymyRvSwzGDhlSPWlzx8vyyza6M2a-MUghJGe6-lsXEw__&Key-Pair-Id=K3EI6M078Z3AC3","height":1440,"width":2159},"inferredFromUrl":false},"qa_pairs":{"kind":"list like","value":{"question":["How many unique income levels are represented in the chart?","Which income level has the highest economic impact for households with four family members and structural damage?","What type of property damage is associated with the lowest economic impact in the chart?","How does the economic impact of major cracks for High-income families compare to that for Low-income families?","What is the size of the bubble representing low-income families with three members and structural damage?","What is the total economic impact for households with two family members across all income levels?","Which type of property damage has the highest frequency for High-income households?","Is the economic impact generally higher for households with more members regardless of income level?","What is the difference in economic impact between structural damage for Middle-income families with two members and Major cracks for the same group with four members?","How does the number of households with structural damage compare between Low and Middle-income groups?"],"explanation":["The chart uses different colors and markers to represent different income levels. By looking at the legend, we can see three categories: Low, Middle, and High Income.","According to the annotations on the plot, the highest economic impact for households with four family members and structural damage comes from the Middle-income group. This can be determined by comparing the y-axis values for the relevant data points.","The type of property damage with the lowest economic impact is identified by locating the bubble with the lowest position on the y-axis. Minor Cracks in the Low-income group have an economic impact of 0.7 million USD.","For High-income families, major cracks have an economic impact of 2 million USD. For Low-income families, the economic impact of major cracks is 1.6 million USD. By comparing these values, we see that the impact is greater for High-income families.","The size of the bubble is proportional to the number of households. For low-income families with three members and structural damage, there are 50 households. The chart multiplies the number by 5 for visibility, so the bubble size represents 250 units.","Sum the economic impact for all income levels with two family members. For Low-income: 1.2M; for Middle-income: 1.2M; for High-income: 0.9M. The total is 1.2 + 1.2 + 0.9 = 3.3 million USD.","The size of each bubble points to the number of households. The bubble for Minor Cracks in High-income households with four members is the largest, suggesting the highest frequency for property damage.","By comparing bubbles across the chart, we generally see that bubbles representing households with more family members are positioned higher on the y-axis, indicating a higher economic impact.","For Middle-income families, structural damage for two members is 1.2M, and major cracks for four members is 1.8M. The difference is 1.8 - 1.2 = 0.6 million USD.","The size of the bubbles indicates the number of households. Low-income has 50 (3 members) + 30 (2 members) = 80 households. Middle-income has 40 (4 members) + 20 (2 members) = 60 households. Therefore, Low-income families have more households with structural damage."],"answer":["3","Middle","Minor Cracks","Higher for High-income families","250 units","3.3 million USD","Minor Cracks","Yes","0.6 million USD","More in Low-income"]},"string":"{\n \"question\": [\n \"How many unique income levels are represented in the chart?\",\n \"Which income level has the highest economic impact for households with four family members and structural damage?\",\n \"What type of property damage is associated with the lowest economic impact in the chart?\",\n \"How does the economic impact of major cracks for High-income families compare to that for Low-income families?\",\n \"What is the size of the bubble representing low-income families with three members and structural damage?\",\n \"What is the total economic impact for households with two family members across all income levels?\",\n \"Which type of property damage has the highest frequency for High-income households?\",\n \"Is the economic impact generally higher for households with more members regardless of income level?\",\n \"What is the difference in economic impact between structural damage for Middle-income families with two members and Major cracks for the same group with four members?\",\n \"How does the number of households with structural damage compare between Low and Middle-income groups?\"\n ],\n \"explanation\": [\n \"The chart uses different colors and markers to represent different income levels. By looking at the legend, we can see three categories: Low, Middle, and High Income.\",\n \"According to the annotations on the plot, the highest economic impact for households with four family members and structural damage comes from the Middle-income group. This can be determined by comparing the y-axis values for the relevant data points.\",\n \"The type of property damage with the lowest economic impact is identified by locating the bubble with the lowest position on the y-axis. Minor Cracks in the Low-income group have an economic impact of 0.7 million USD.\",\n \"For High-income families, major cracks have an economic impact of 2 million USD. For Low-income families, the economic impact of major cracks is 1.6 million USD. By comparing these values, we see that the impact is greater for High-income families.\",\n \"The size of the bubble is proportional to the number of households. For low-income families with three members and structural damage, there are 50 households. The chart multiplies the number by 5 for visibility, so the bubble size represents 250 units.\",\n \"Sum the economic impact for all income levels with two family members. For Low-income: 1.2M; for Middle-income: 1.2M; for High-income: 0.9M. The total is 1.2 + 1.2 + 0.9 = 3.3 million USD.\",\n \"The size of each bubble points to the number of households. The bubble for Minor Cracks in High-income households with four members is the largest, suggesting the highest frequency for property damage.\",\n \"By comparing bubbles across the chart, we generally see that bubbles representing households with more family members are positioned higher on the y-axis, indicating a higher economic impact.\",\n \"For Middle-income families, structural damage for two members is 1.2M, and major cracks for four members is 1.8M. The difference is 1.8 - 1.2 = 0.6 million USD.\",\n \"The size of the bubbles indicates the number of households. Low-income has 50 (3 members) + 30 (2 members) = 80 households. Middle-income has 40 (4 members) + 20 (2 members) = 60 households. Therefore, Low-income families have more households with structural damage.\"\n ],\n \"answer\": [\n \"3\",\n \"Middle\",\n \"Minor Cracks\",\n \"Higher for High-income families\",\n \"250 units\",\n \"3.3 million USD\",\n \"Minor Cracks\",\n \"Yes\",\n \"0.6 million USD\",\n \"More in Low-income\"\n ]\n}"},"metadata":{"kind":"string","value":"{\n \"figure_type\": \"Bubble Chart\",\n \"persona\": \"A resident who experienced property damage due to a recent earthquake and supports the government official's efforts\",\n \"topic\": \"Demographics and economic impacts of the earthquake on affected households, categorized by income level, number of family members, and type of property damage\"\n}"},"data":{"kind":"string","value":"Income Level,Number of Family Members,Type of Property Damage,Number of Households,Economic Impact (USD)\nLow,3,Structural Damage,50,2000000\nLow,2,Structural Damage,30,1200000\nLow,4,Minor Cracks,70,700000\nLow,3,Major Cracks,40,1600000\nMiddle,4,Structural Damage,40,2400000\nMiddle,3,Minor Cracks,60,900000\nMiddle,2,Structural Damage,20,1200000\nMiddle,4,Major Cracks,30,1800000\nHigh,5,Structural Damage,20,3000000\nHigh,4,Minor Cracks,50,1500000\nHigh,3,Major Cracks,25,2000000\nHigh,2,Minor Cracks,30,900000"},"code":{"kind":"string","value":"import numpy as np\nimport matplotlib.pyplot as plt\nfrom io import BytesIO\nfrom PIL import Image\n\ndef generate_plot(df):\n plt.style.use('grayscale')\n \n fig, ax = plt.subplots(figsize=(12, 8))\n \n income_levels = df['Income Level'].unique()\n colors = ['#1f77b4', '#ff7f0e', '#2ca02c']\n markers = ['o', 's', '^']\n \n for i, income in enumerate(income_levels):\n data = df[df['Income Level'] == income]\n x = data['Number of Family Members']\n y = data['Economic Impact (USD)'] / 1e6 # Convert to millions\n size = data['Number of Households'] * 5 # Adjust size for visibility\n \n scatter = ax.scatter(x, y, s=size, c=[colors[i]], marker=markers[i], alpha=0.7, \n label=f'{income} Income')\n \n # Annotate data points with damage type\n for j, txt in enumerate(data['Type of Property Damage']):\n ax.annotate(txt, (x.iloc[j], y.iloc[j]), xytext=(5, 5), \n textcoords='offset points', fontsize=8, alpha=0.8)\n \n ax.set_xlabel('Number of Family Members', fontsize=12, fontweight='bold')\n ax.set_ylabel('Economic Impact (Millions USD)', fontsize=12, fontweight='bold')\n ax.set_title('Economic Impact of Earthquake by Income Level, Family Size, and Damage Type', \n fontsize=14, fontweight='bold')\n \n ax.grid(True, linestyle='--', alpha=0.7)\n ax.set_axisbelow(True)\n \n # Customize tick labels\n ax.tick_params(axis='both', which='major', labelsize=10)\n \n # Add legend\n legend = ax.legend(title='Income Level', fontsize=10, title_fontsize=12)\n legend.get_frame().set_alpha(0.8)\n \n # Add text about bubble size\n ax.text(0.95, 0.05, 'Bubble size represents\\nnumber of households', \n transform=ax.transAxes, fontsize=10, ha='right', va='bottom', \n bbox=dict(boxstyle='round', facecolor='white', edgecolor='gray', alpha=0.8))\n \n plt.tight_layout()\n \n # Save plot to BytesIO object\n buf = BytesIO()\n plt.savefig(buf, format='png', dpi=300, bbox_inches='tight')\n buf.seek(0)\n \n # Convert to PIL Image\n img = Image.open(buf)\n \n return img"}}},{"rowIdx":11,"cells":{"id":{"kind":"string","value":"PlotlyChartPipeline_bar_claude_1-748"},"image":{"kind":"image","value":{"src":"https://datasets-server.huggingface.co/assets/allenai/CoSyn-400K/--/86e46e1fd5e754d056169f0fb38f06c6997ff7de/--/chart/train/11/image/image.jpg?Expires=1741922929&Signature=aiqn8SZQHtQi3vgaXDayJp6qxWuwneN6m454S2DQkU6cYfPaJbXmZWlkCrqUt6pNd9qdy2BljsZfpkww-RHIgRbBeuwXhJGo9~k-2mhfMEnkM7BzpF-xpeTiuHPM6CKP078fFEnjuuUadDwFrA5Fxnj-2AnGpSUDeKU706cF1wdoG6LDV3gNPG5-rpJJKW~GUACQq~orGM6yn1yJ-pkG6brOSH0zFAOfL4-HCuUy0GaMTKV1Vrr7GIdpL2ZSCiDkshgQsoLHjxjKJ-KZxkG2-6XWIAy3~BXq8KpafbYdvJ0t--sI70NUHW8rNJmVG7D3x0f9fuam~VZ06y0KxwuTJw__&Key-Pair-Id=K3EI6M078Z3AC3","height":1200,"width":2000},"inferredFromUrl":false},"qa_pairs":{"kind":"list like","value":{"question":["What is the competition in which Seán Carrabine scored the highest number of goals?","How many goals did Seán Carrabine score in total in the Under-19 National League and Under-21 National League combined?","Which competition has a higher number of goals scored by Seán Carrabine: the FAI Cup or the Friendly Matches?","What is the cumulative number of goals after the FAI Cup in the sorted order?","In which competition did Seán Carrabine score twice and have a cumulative goal count of 18 before the start of that competition?","What is the cumulative number of goals at the end of the competition list?","Between Leinster Senior Cup and President's Cup, which competition has fewer goals and by how many?","What is the color used for the cumulative goals line, and what marker symbolizes it?"],"explanation":["The competition with the highest bar represents the highest number of goals. The tallest bar belongs to the \"League of Ireland Premier Division\" with 7 goals.","Seán Carrabine scored 6 goals in the Under-19 National League and 5 goals in the Under-21 National League. Adding these together gives 6 + 5 = 11 goals.","By comparing the height of the bars for the FAI Cup and the Friendly Matches, we see that the Friendly Matches bar is taller with 4 goals, whereas the FAI Cup has 3 goals.","The cumulative number of goals for each competition is plotted using a golden line. The order after sorting starts with the League of Ireland Premier Division. Summing up the goals until the FAI Cup: (7 for League of Ireland Premier Division + 6 for Under-19 National League + 5 for Under-21 National League + 4 for Friendly Matches + 3 for FAI Cup) gives 25.","The bar plot shows that Seán Carrabine scored 2 goals in the League of Ireland Cup and Leinster Senior Cup. The cumulative goal line reaches 18 after adding the goals for Under-21 National League (5), Under-19 National League (6), and League of Ireland Premier Division (7). Thus, the next competition where he scored 2 goals is the League of Ireland Cup.","The cumulative goals line reaches its highest value at the end of the list. Adding all goals across the competitions: (7 + 6 + 5 + 4 + 3 + 2 + 2 + 1 + 1 + 1) gives a total of 32.","Comparing the heights of the bars for Leinster Senior Cup and President's Cup shows that the President's Cup has 1 goal, while the Leinster Senior Cup has 2 goals. The difference is 2 - 1 = 1 goal.","The cumulative goals are represented by a golden line with star-shaped markers as observed from the legend and the plot."],"answer":["League of Ireland Premier Division","11","Friendly Matches","25","League of Ireland Cup","32","President's Cup, 1 goal","Golden, star markers"]},"string":"{\n \"question\": [\n \"What is the competition in which Seán Carrabine scored the highest number of goals?\",\n \"How many goals did Seán Carrabine score in total in the Under-19 National League and Under-21 National League combined?\",\n \"Which competition has a higher number of goals scored by Seán Carrabine: the FAI Cup or the Friendly Matches?\",\n \"What is the cumulative number of goals after the FAI Cup in the sorted order?\",\n \"In which competition did Seán Carrabine score twice and have a cumulative goal count of 18 before the start of that competition?\",\n \"What is the cumulative number of goals at the end of the competition list?\",\n \"Between Leinster Senior Cup and President's Cup, which competition has fewer goals and by how many?\",\n \"What is the color used for the cumulative goals line, and what marker symbolizes it?\"\n ],\n \"explanation\": [\n \"The competition with the highest bar represents the highest number of goals. The tallest bar belongs to the \\\"League of Ireland Premier Division\\\" with 7 goals.\",\n \"Seán Carrabine scored 6 goals in the Under-19 National League and 5 goals in the Under-21 National League. Adding these together gives 6 + 5 = 11 goals.\",\n \"By comparing the height of the bars for the FAI Cup and the Friendly Matches, we see that the Friendly Matches bar is taller with 4 goals, whereas the FAI Cup has 3 goals.\",\n \"The cumulative number of goals for each competition is plotted using a golden line. The order after sorting starts with the League of Ireland Premier Division. Summing up the goals until the FAI Cup: (7 for League of Ireland Premier Division + 6 for Under-19 National League + 5 for Under-21 National League + 4 for Friendly Matches + 3 for FAI Cup) gives 25.\",\n \"The bar plot shows that Seán Carrabine scored 2 goals in the League of Ireland Cup and Leinster Senior Cup. The cumulative goal line reaches 18 after adding the goals for Under-21 National League (5), Under-19 National League (6), and League of Ireland Premier Division (7). Thus, the next competition where he scored 2 goals is the League of Ireland Cup.\",\n \"The cumulative goals line reaches its highest value at the end of the list. Adding all goals across the competitions: (7 + 6 + 5 + 4 + 3 + 2 + 2 + 1 + 1 + 1) gives a total of 32.\",\n \"Comparing the heights of the bars for Leinster Senior Cup and President's Cup shows that the President's Cup has 1 goal, while the Leinster Senior Cup has 2 goals. The difference is 2 - 1 = 1 goal.\",\n \"The cumulative goals are represented by a golden line with star-shaped markers as observed from the legend and the plot.\"\n ],\n \"answer\": [\n \"League of Ireland Premier Division\",\n \"11\",\n \"Friendly Matches\",\n \"25\",\n \"League of Ireland Cup\",\n \"32\",\n \"President's Cup, 1 goal\",\n \"Golden, star markers\"\n ]\n}"},"metadata":{"kind":"string","value":"{\n \"figure_type\": \"bar chart\",\n \"persona\": \"A hardcore Seán Carrabine fan and an ardent follower of Irish football\",\n \"topic\": \"Seán Carrabine's goal-scoring statistics across different competitions\"\n}"},"data":{"kind":"string","value":"Competition,Goals\nLeague of Ireland Premier Division,7\nFAI Cup,3\nLeague of Ireland Cup,2\nUEFA Europa Conference League Qualifying,1\nLeinster Senior Cup,2\nPresident's Cup,1\nFriendly Matches,4\nUnder-21 National League,5\nUnder-19 National League,6\nFAI Intermediate Cup,1"},"code":{"kind":"string","value":"import numpy as np\nimport plotly.graph_objects as go\nfrom plotly.subplots import make_subplots\nfrom io import BytesIO\nfrom PIL import Image\n\ndef generate_plot(data):\n # Sort data by goals in descending order\n data_sorted = data.sort_values('Goals', ascending=False)\n\n # Create figure with secondary y-axis\n fig = make_subplots(specs=[[{\"secondary_y\": True}]])\n\n # Add bars\n fig.add_trace(\n go.Bar(\n x=data_sorted['Competition'],\n y=data_sorted['Goals'],\n name='Goals',\n marker_color='rgba(0, 128, 0, 0.8)',\n marker_line_color='rgba(0, 128, 0, 1.0)',\n marker_line_width=1.5,\n textposition='outside',\n texttemplate='%{y}',\n textfont=dict(size=12, color='black')\n )\n )\n\n # Add line for cumulative goals\n cumulative_goals = np.cumsum(data_sorted['Goals'])\n fig.add_trace(\n go.Scatter(\n x=data_sorted['Competition'],\n y=cumulative_goals,\n name='Cumulative Goals',\n line=dict(color='gold', width=3),\n mode='lines+markers',\n marker=dict(size=8, symbol='star', line=dict(width=2, color='darkgoldenrod'))\n ),\n secondary_y=True\n )\n\n # Customize layout\n fig.update_layout(\n title={\n 'text': \"Seán Carrabine's Goal-Scoring Statistics\",\n 'font': {'size': 24, 'color': 'darkgreen', 'family': 'Arial Black'},\n 'x': 0.5,\n 'y': 0.95\n },\n xaxis={\n 'title': 'Competition',\n 'tickangle': 45,\n 'tickfont': {'size': 10},\n 'title_font': {'size': 16, 'color': 'darkgreen'}\n },\n yaxis={\n 'title': 'Goals',\n 'title_font': {'size': 16, 'color': 'darkgreen'},\n 'tickfont': {'size': 12},\n 'showgrid': True,\n 'gridcolor': 'lightgray'\n },\n yaxis2={\n 'title': 'Cumulative Goals',\n 'title_font': {'size': 16, 'color': 'darkgoldenrod'},\n 'tickfont': {'size': 12},\n 'overlaying': 'y',\n 'side': 'right'\n },\n legend={\n 'x': 1.05,\n 'y': 1,\n 'bgcolor': 'rgba(255, 255, 255, 0.8)',\n 'bordercolor': 'rgba(0, 0, 0, 0.3)',\n 'borderwidth': 1\n },\n font=dict(family='Arial'),\n plot_bgcolor='white',\n width=1000,\n height=600,\n margin=dict(l=50, r=50, t=80, b=100)\n )\n\n # Add a watermark\n fig.add_annotation(\n text=\"Seán Carrabine Fan Club\",\n x=0.5,\n y=0.5,\n xref=\"paper\",\n yref=\"paper\",\n showarrow=False,\n font=dict(size=40, color=\"rgba(0,100,0,0.1)\"),\n textangle=-30\n )\n\n # Save plot to BytesIO object\n img_bytes = BytesIO()\n fig.write_image(img_bytes, format=\"png\", scale=2)\n img_bytes.seek(0)\n \n # Return plot as PIL Image object\n return Image.open(img_bytes)"}}},{"rowIdx":12,"cells":{"id":{"kind":"string","value":"MatplotlibChartPipeline_stair_1-31"},"image":{"kind":"image","value":{"src":"https://datasets-server.huggingface.co/assets/allenai/CoSyn-400K/--/86e46e1fd5e754d056169f0fb38f06c6997ff7de/--/chart/train/12/image/image.jpg?Expires=1741922929&Signature=t7s6Y8kHz0eZm5mAI6Dyt-FX4JLzvsaNj5-Lue4MoLFbDWklUAVhgLokM28XmpmtDBLZjoBgPsxXLvPdsqubv1kp~HlQzYhwpuMuhJ-2oGVSI3IY92Iz-vt4F3qcg9bfwGk9FWZ9X2gon1eRn7sSpT04z6j9330oTTA3zcuIqOddAzBFiDRNhOptH2Gv2yMd2ZvG125oFkraCR1kJ25Gb107k~0YJAtak1k4DizqSGJfhdiTTT9GxWuHDEwZzyGdF3DYou~Kve5ZWza2MmA~Nw0PAatoPM1yilU5rpK89TYWf6EX9p0bkhZbqRKloSKXfXgWG-ccHrSDBDhickGiuQ__&Key-Pair-Id=K3EI6M078Z3AC3","height":1440,"width":2157},"inferredFromUrl":false},"qa_pairs":{"kind":"list like","value":{"question":["What is the title of the figure?","Which injury type had the highest frequency in 2018 for pitchers?","What is the frequency difference for hamstring strains in 2018 between outfielders and catchers?","Which player position had the lowest frequency of knee injuries in 2020?","By how much did shoulder injuries for infielders decrease from 2018 to 2020?","Did the frequency of hamstring strains for outfielders increase or decrease from 2018 to 2019?","Identify the year with the highest total frequency of injuries for catchers.","In 2020, what was the total frequency of shoulder injuries across all player positions?","Which injury type experienced the most significant reduction in frequency for outfielders from 2018 to 2020?"],"explanation":["The title is written at the top of the figure and usually provides a summary of what the plot is about. The title in this figure reads 'Frequency of Common Injuries Over Time by Player Position'.","To determine this, look at the highest point on the y-axis for pitchers (marked in the legend) over the year 2018. The data shows the highest value for shoulder injury in 2018 for pitchers.","First, find the frequency values for hamstring strains for both outfielders and catchers in 2018. The frequencies are 24 for outfielders and 15 for catchers. The difference is 24 - 15.","Locate the knee injury data for 2020 and compare the values for each position. Pitchers have the lowest frequency at 5.","Find the frequency of shoulder injuries for infielders in 2018 and 2020. The values are 25 in 2018 and 8 in 2020. Calculate the decrease by subtracting 8 from 25.","Compare the frequencies for 2018 and 2019. In 2018 it was 24 and in 2019 it was 20. Since 20 is less than 24, the frequency decreased.","To determine this, sum the frequencies for each injury type across all years (2018, 2019, 2020) for catchers. The totals are highest in 2018.","Sum the frequencies of shoulder injuries for all player positions (Outfielder, Infielder, Pitcher, Catcher) for the year 2020. The values are 12 + 8 + 18 + 10, giving a total of 48.","Calculate the differences in frequency for each injury type (hamstring strain, shoulder injury, knee injury) for outfielders from 2018 to 2020. The largest reduction is for shoulder injuries, from 30 to 12."],"answer":["Frequency of Common Injuries Over Time by Player Position","Shoulder Injury","9","Pitcher","17","Decrease","2018","48","Shoulder Injury"]},"string":"{\n \"question\": [\n \"What is the title of the figure?\",\n \"Which injury type had the highest frequency in 2018 for pitchers?\",\n \"What is the frequency difference for hamstring strains in 2018 between outfielders and catchers?\",\n \"Which player position had the lowest frequency of knee injuries in 2020?\",\n \"By how much did shoulder injuries for infielders decrease from 2018 to 2020?\",\n \"Did the frequency of hamstring strains for outfielders increase or decrease from 2018 to 2019?\",\n \"Identify the year with the highest total frequency of injuries for catchers.\",\n \"In 2020, what was the total frequency of shoulder injuries across all player positions?\",\n \"Which injury type experienced the most significant reduction in frequency for outfielders from 2018 to 2020?\"\n ],\n \"explanation\": [\n \"The title is written at the top of the figure and usually provides a summary of what the plot is about. The title in this figure reads 'Frequency of Common Injuries Over Time by Player Position'.\",\n \"To determine this, look at the highest point on the y-axis for pitchers (marked in the legend) over the year 2018. The data shows the highest value for shoulder injury in 2018 for pitchers.\",\n \"First, find the frequency values for hamstring strains for both outfielders and catchers in 2018. The frequencies are 24 for outfielders and 15 for catchers. The difference is 24 - 15.\",\n \"Locate the knee injury data for 2020 and compare the values for each position. Pitchers have the lowest frequency at 5.\",\n \"Find the frequency of shoulder injuries for infielders in 2018 and 2020. The values are 25 in 2018 and 8 in 2020. Calculate the decrease by subtracting 8 from 25.\",\n \"Compare the frequencies for 2018 and 2019. In 2018 it was 24 and in 2019 it was 20. Since 20 is less than 24, the frequency decreased.\",\n \"To determine this, sum the frequencies for each injury type across all years (2018, 2019, 2020) for catchers. The totals are highest in 2018.\",\n \"Sum the frequencies of shoulder injuries for all player positions (Outfielder, Infielder, Pitcher, Catcher) for the year 2020. The values are 12 + 8 + 18 + 10, giving a total of 48.\",\n \"Calculate the differences in frequency for each injury type (hamstring strain, shoulder injury, knee injury) for outfielders from 2018 to 2020. The largest reduction is for shoulder injuries, from 30 to 12.\"\n ],\n \"answer\": [\n \"Frequency of Common Injuries Over Time by Player Position\",\n \"Shoulder Injury\",\n \"9\",\n \"Pitcher\",\n \"17\",\n \"Decrease\",\n \"2018\",\n \"48\",\n \"Shoulder Injury\"\n ]\n}"},"metadata":{"kind":"string","value":"{\n \"figure_type\": \"stair plot\",\n \"persona\": \"An student applying for a baseball coach job\",\n \"topic\": \"Frequency of Common Injuries Over Time with Breakdown by Training Intensity and Player Position\"\n}"},"data":{"kind":"string","value":"Injury Type,Year,Position,Training Intensity,Frequency\nHamstring Strain,2018,Outfielder,High,24\nHamstring Strain,2018,Infielder,High,18\nHamstring Strain,2018,Pitcher,High,20\nHamstring Strain,2018,Catcher,High,15\nHamstring Strain,2019,Outfielder,Medium,20\nHamstring Strain,2019,Infielder,Medium,15\nHamstring Strain,2019,Pitcher,Medium,22\nHamstring Strain,2019,Catcher,Medium,12\nHamstring Strain,2020,Outfielder,Low,10\nHamstring Strain,2020,Infielder,Low,9\nHamstring Strain,2020,Pitcher,Low,12\nHamstring Strain,2020,Catcher,Low,8\nShoulder Injury,2018,Outfielder,High,30\nShoulder Injury,2018,Infielder,High,25\nShoulder Injury,2018,Pitcher,High,35\nShoulder Injury,2018,Catcher,High,28\nShoulder Injury,2019,Outfielder,Medium,22\nShoulder Injury,2019,Infielder,Medium,19\nShoulder Injury,2019,Pitcher,Medium,30\nShoulder Injury,2019,Catcher,Medium,24\nShoulder Injury,2020,Outfielder,Low,12\nShoulder Injury,2020,Infielder,Low,8\nShoulder Injury,2020,Pitcher,Low,18\nShoulder Injury,2020,Catcher,Low,10\nKnee Injury,2018,Outfielder,High,18\nKnee Injury,2018,Infielder,High,16\nKnee Injury,2018,Pitcher,High,14\nKnee Injury,2018,Catcher,High,20\nKnee Injury,2019,Outfielder,Medium,12\nKnee Injury,2019,Infielder,Medium,15\nKnee Injury,2019,Pitcher,Medium,11\nKnee Injury,2019,Catcher,Medium,16\nKnee Injury,2020,Outfielder,Low,6\nKnee Injury,2020,Infielder,Low,7\nKnee Injury,2020,Pitcher,Low,5\nKnee Injury,2020,Catcher,Low,9"},"code":{"kind":"string","value":"import numpy as np\nimport matplotlib.pyplot as plt\nfrom io import BytesIO\nfrom PIL import Image\n\ndef generate_plot(data):\n plt.style.use('default')\n \n fig, ax = plt.subplots(figsize=(12, 8))\n \n injury_types = data['Injury Type'].unique()\n positions = data['Position'].unique()\n colors = plt.cm.viridis(np.linspace(0, 1, len(positions)))\n \n for i, injury in enumerate(injury_types):\n injury_data = data[data['Injury Type'] == injury]\n \n for j, position in enumerate(positions):\n pos_data = injury_data[injury_data['Position'] == position]\n ax.step(pos_data['Year'], pos_data['Frequency'], where='mid', \n label=f\"{injury} - {position}\", \n color=colors[j], \n linewidth=2.5, \n alpha=0.8, \n marker='o', \n markersize=8)\n \n ax.set_xlabel('Year', fontsize=14, fontweight='bold')\n ax.set_ylabel('Frequency', fontsize=14, fontweight='bold')\n ax.set_title('Frequency of Common Injuries Over Time\\nby Player Position', \n fontsize=18, fontweight='bold', pad=20)\n \n ax.grid(True, linestyle='--', alpha=0.7)\n ax.set_xticks(data['Year'].unique())\n ax.tick_params(axis='both', which='major', labelsize=12)\n \n plt.legend(bbox_to_anchor=(1.05, 1), loc='upper left', fontsize=10)\n \n plt.tight_layout()\n \n buffer = BytesIO()\n plt.savefig(buffer, format='png', dpi=300, bbox_inches='tight')\n buffer.seek(0)\n image = Image.open(buffer)\n \n return image"}}},{"rowIdx":13,"cells":{"id":{"kind":"string","value":"HTMLChartPipeline_pie_5-250"},"image":{"kind":"image","value":{"src":"https://datasets-server.huggingface.co/assets/allenai/CoSyn-400K/--/86e46e1fd5e754d056169f0fb38f06c6997ff7de/--/chart/train/13/image/image.jpg?Expires=1741922929&Signature=hF0pIOZ8zelPzetkTWlbKF9CXEIrSYEDKoIj9xFS99AqHs5O6Bt6g97PgKT4OODY8kBsXMmedpOBz-nZXQRM7X13R-5a9KSbTU6BVdHc2WZTNDHdIUxYAU0J1187FpVwfU570QNYaQig~pjeYqgHGNWn3igCQPCy9nps~hhks2fBBbnb7nbt04hkO8SnOJjC1w0olAuE-5dxV3Xpy3YDaqvNfjojguMvLvtKUR0kU5EopdCCMEM9ZhLRigjWPtGAhGlw5H~Huio~Usa9p6L3Ncnm16ecdZjciT9AMxs4CgT071HhSa3aLe5i3llRaLQxYc9F4dfvnQM--JWIF1g1aw__&Key-Pair-Id=K3EI6M078Z3AC3","height":800,"width":661},"inferredFromUrl":false},"qa_pairs":{"kind":"list like","value":{"question":["What percentage of fans prefer digital artwork?","Which is more popular, animations or traditional artwork?","What is the combined percentage of fans who prefer manga and merchandise?","Which content type is the least preferred?","How much more popular is manga compared to traditional artwork?","What is the second most preferred content type?","Which content types have a percentage that is above 20%?","What is the average preference percentage for all content types?","By what factor is manga more popular than merchandise?"],"explanation":["Look at the segment labeled \"デジタルアート\" in the pie chart, which represents the digital artwork preference. The chart indicates that 25% of fans prefer digital artwork.","Compare the percentages of the segments labeled \"アニメーション\" and \"伝統的なアート\". Animations have a percentage of 20%, while traditional artwork has 15%. Therefore, animations are more popular.","Add the percentages of the segments labeled \"マンガ\" and \"グッズ\". Manga has 30% and merchandise has 10%. Therefore, 30% + 10% = 40%.","Identify the segment with the smallest percentage. The segment labeled \"グッズ\" (merchandise) has the lowest percentage at 10%.","Subtract the percentage of traditional artwork from manga. Manga has 30% and traditional artwork has 15%. Therefore, 30% - 15% = 15%.","Look at the pie chart segments in descending order of their percentages. The first is manga (30%), and the second highest is digital artwork (25%).","Identify segments with percentages greater than 20%. Manga has 30% and digital artwork has 25%, both are above 20%.","Sum the percentages of all content types and divide by the number of content types. The percentages are 30, 25, 20, 15, and 10. Summing them gives 100. Dividing by 5 gives 100 / 5 = 20%.","Divide the percentage of manga by the percentage of merchandise. Manga has 30% and merchandise has 10%, so 30 / 10 = 3."],"answer":["25%","Animations","40%","Merchandise","15%","Digital Artwork","Manga, Digital Artwork","20%","3"]},"string":"{\n \"question\": [\n \"What percentage of fans prefer digital artwork?\",\n \"Which is more popular, animations or traditional artwork?\",\n \"What is the combined percentage of fans who prefer manga and merchandise?\",\n \"Which content type is the least preferred?\",\n \"How much more popular is manga compared to traditional artwork?\",\n \"What is the second most preferred content type?\",\n \"Which content types have a percentage that is above 20%?\",\n \"What is the average preference percentage for all content types?\",\n \"By what factor is manga more popular than merchandise?\"\n ],\n \"explanation\": [\n \"Look at the segment labeled \\\"デジタルアート\\\" in the pie chart, which represents the digital artwork preference. The chart indicates that 25% of fans prefer digital artwork.\",\n \"Compare the percentages of the segments labeled \\\"アニメーション\\\" and \\\"伝統的なアート\\\". Animations have a percentage of 20%, while traditional artwork has 15%. Therefore, animations are more popular.\",\n \"Add the percentages of the segments labeled \\\"マンガ\\\" and \\\"グッズ\\\". Manga has 30% and merchandise has 10%. Therefore, 30% + 10% = 40%.\",\n \"Identify the segment with the smallest percentage. The segment labeled \\\"グッズ\\\" (merchandise) has the lowest percentage at 10%.\",\n \"Subtract the percentage of traditional artwork from manga. Manga has 30% and traditional artwork has 15%. Therefore, 30% - 15% = 15%.\",\n \"Look at the pie chart segments in descending order of their percentages. The first is manga (30%), and the second highest is digital artwork (25%).\",\n \"Identify segments with percentages greater than 20%. Manga has 30% and digital artwork has 25%, both are above 20%.\",\n \"Sum the percentages of all content types and divide by the number of content types. The percentages are 30, 25, 20, 15, and 10. Summing them gives 100. Dividing by 5 gives 100 / 5 = 20%.\",\n \"Divide the percentage of manga by the percentage of merchandise. Manga has 30% and merchandise has 10%, so 30 / 10 = 3.\"\n ],\n \"answer\": [\n \"25%\",\n \"Animations\",\n \"40%\",\n \"Merchandise\",\n \"15%\",\n \"Digital Artwork\",\n \"Manga, Digital Artwork\",\n \"20%\",\n \"3\"\n ]\n}"},"metadata":{"kind":"string","value":"{\n \"figure_type\": \"Pie chart with five data point\",\n \"persona\": \"a content creator and illustrator from Japan who has been heavily impacted by piracy\",\n \"topic\": \"Preferred Content Types by Fans of Japanese Illustrators (e.g., manga, digital artwork, animations, traditional artwork, merchandise)\"\n}"},"data":{"kind":"string","value":"{\n \"PreferredContentTypesByFans\": [\n {\n \"ContentType\": \"Manga\",\n \"Percentage\": 30\n },\n {\n \"ContentType\": \"Digital Artwork\",\n \"Percentage\": 25\n },\n {\n \"ContentType\": \"Animations\",\n \"Percentage\": 20\n },\n {\n \"ContentType\": \"Traditional Artwork\",\n \"Percentage\": 15\n },\n {\n \"ContentType\": \"Merchandise\",\n \"Percentage\": 10\n }\n ]\n}"},"code":{"kind":"string","value":"\n\n\n \n \n \n \n \n\n\n \n\n \n\n"}}},{"rowIdx":14,"cells":{"id":{"kind":"string","value":"MatplotlibChartPipeline_error_1-218"},"image":{"kind":"image","value":{"src":"https://datasets-server.huggingface.co/assets/allenai/CoSyn-400K/--/86e46e1fd5e754d056169f0fb38f06c6997ff7de/--/chart/train/14/image/image.jpg?Expires=1741922929&Signature=gXbEzrfUkvHG99rogATuwDc8IzMs6Jsx8x~23Hkp1Kq~L2LUwFYMbs23rKTmiVarp8e2tk-Kq7Z7c1zu94igh-8DYFyRWR5Mz5XgYz4Qsp6sQmCosbQDJplq5zqkOGcD0bHzK-3TBF51TNKP2JrPtWpyt~I25SKQk4k-Aus~m3xYuRLoeUROgWb895cX~i2SoLcJeMTNP48IEFxCXwf0JHHSsyIt9SFrjE0OYOQfREFsd1wnuEqtVV4ye7hNducOISZmEt6RXIqoEYmeo39b32QetNPvl~xkT9dbHjRlTWeurIEWGZeB7eX~1-1-qYTR5HXxfOlHFwgZ8zBvUWSvOQ__&Key-Pair-Id=K3EI6M078Z3AC3","height":1440,"width":2169},"inferredFromUrl":false},"qa_pairs":{"kind":"list like","value":{"question":["What is the title of the figure?","Which age group has the highest popularity for Smart Lighting?","What is the installation error rate for Smart Security Cameras in the 65+ age group?","For the age group 45-54, what is the difference in popularity between Smart Thermostats and Smart Security Cameras?","Which device category shows the largest variation in popularity across the age groups?","Comparing the age group 35-44, which device has the lowest popularity rate?","What is the average popularity of Smart Thermostats across all age groups?","How does the popularity of Smart Security Cameras change from the 18-24 age group to the 65+ age group?","In which age group do installation error rates for Smart Lighting stay relatively low?","For the age group 55-64, calculate the sum of installation error rates for all devices."],"explanation":["The title is generally placed at the top of the figure, often in bold to make it prominent. Here, the title \"Smart Home Device Popularity and Installation Error Rates\" is presented at the top in bold.","To answer this, look for the highest bar in the Smart Lighting category. The age group 25-34 has the highest popularity for Smart Lighting at 80%.","Installation error rates are marked by the error bars on top of the bars. For the 65+ age group and Smart Security Cameras, the error bar extends to 30%.","Look at the bars for Smart Thermostats and Smart Security Cameras in the 45-54 age group. The popularity of Smart Thermostats is 55%, and for Smart Security Cameras, it is 45%. Subtract the smaller value from the larger one: 55% - 45% = 10%.","Identify the device category with the widest range between its highest and lowest values. Smart Lighting ranges from 50% (65+) to 80% (25-34), making it the device with the largest variation in popularity.","Check the heights of the bars for each device category in the 35-44 age group. Smart Security Cameras have the lowest popularity at 50%.","Sum the popularity percentages for Smart Thermostats in all age groups and divide by the number of age groups (65 + 70 + 60 + 55 + 50 + 40) / 6. Calculate as follows: (65 + 70 + 60 + 55 + 50 + 40) / 6 = 340/6 = 56.67%.","Note the popularity percentages for Smart Security Cameras in the 18-24 age group (55%) and compare it with the 65+ age group (35%). The popularity decreases from 55% to 35%.","Look at the heights of the error bars for Smart Lighting in each age group. The error rate for Smart Lighting in the 25-34 age group is 8%, which is relatively low compared to other groups.","Add the installation error rates for Smart Thermostats (16%), Smart Lighting (15%), and Smart Security Cameras (28%) in the 55-64 age group. Calculate as follows: 16% + 15% + 28% = 59%."],"answer":["Smart Home Device Popularity and Installation Error Rates","25-34","30%","10%","Smart Lighting","Smart Security Cameras","56.67%","Decreases by 20%","25-34","59%"]},"string":"{\n \"question\": [\n \"What is the title of the figure?\",\n \"Which age group has the highest popularity for Smart Lighting?\",\n \"What is the installation error rate for Smart Security Cameras in the 65+ age group?\",\n \"For the age group 45-54, what is the difference in popularity between Smart Thermostats and Smart Security Cameras?\",\n \"Which device category shows the largest variation in popularity across the age groups?\",\n \"Comparing the age group 35-44, which device has the lowest popularity rate?\",\n \"What is the average popularity of Smart Thermostats across all age groups?\",\n \"How does the popularity of Smart Security Cameras change from the 18-24 age group to the 65+ age group?\",\n \"In which age group do installation error rates for Smart Lighting stay relatively low?\",\n \"For the age group 55-64, calculate the sum of installation error rates for all devices.\"\n ],\n \"explanation\": [\n \"The title is generally placed at the top of the figure, often in bold to make it prominent. Here, the title \\\"Smart Home Device Popularity and Installation Error Rates\\\" is presented at the top in bold.\",\n \"To answer this, look for the highest bar in the Smart Lighting category. The age group 25-34 has the highest popularity for Smart Lighting at 80%.\",\n \"Installation error rates are marked by the error bars on top of the bars. For the 65+ age group and Smart Security Cameras, the error bar extends to 30%.\",\n \"Look at the bars for Smart Thermostats and Smart Security Cameras in the 45-54 age group. The popularity of Smart Thermostats is 55%, and for Smart Security Cameras, it is 45%. Subtract the smaller value from the larger one: 55% - 45% = 10%.\",\n \"Identify the device category with the widest range between its highest and lowest values. Smart Lighting ranges from 50% (65+) to 80% (25-34), making it the device with the largest variation in popularity.\",\n \"Check the heights of the bars for each device category in the 35-44 age group. Smart Security Cameras have the lowest popularity at 50%.\",\n \"Sum the popularity percentages for Smart Thermostats in all age groups and divide by the number of age groups (65 + 70 + 60 + 55 + 50 + 40) / 6. Calculate as follows: (65 + 70 + 60 + 55 + 50 + 40) / 6 = 340/6 = 56.67%.\",\n \"Note the popularity percentages for Smart Security Cameras in the 18-24 age group (55%) and compare it with the 65+ age group (35%). The popularity decreases from 55% to 35%.\",\n \"Look at the heights of the error bars for Smart Lighting in each age group. The error rate for Smart Lighting in the 25-34 age group is 8%, which is relatively low compared to other groups.\",\n \"Add the installation error rates for Smart Thermostats (16%), Smart Lighting (15%), and Smart Security Cameras (28%) in the 55-64 age group. Calculate as follows: 16% + 15% + 28% = 59%.\"\n ],\n \"answer\": [\n \"Smart Home Device Popularity and Installation Error Rates\",\n \"25-34\",\n \"30%\",\n \"10%\",\n \"Smart Lighting\",\n \"Smart Security Cameras\",\n \"56.67%\",\n \"Decreases by 20%\",\n \"25-34\",\n \"59%\"\n ]\n}"},"metadata":{"kind":"string","value":"{\n \"figure_type\": \"bar chart with error bars\",\n \"persona\": \"An interior designer who incorporates the latest hardware trends into their designs\",\n \"topic\": \"Popularity of Smart Home Devices Across Different Age Groups with Installation Error Rates\"\n}"},"data":{"kind":"string","value":"Age Group,Device Category,Popularity (%),Installation Error Rate (%)\n18-24,Smart Thermostats,65,15\n18-24,Smart Lighting,72,12\n18-24,Smart Security Cameras,55,20\n25-34,Smart Thermostats,70,10\n25-34,Smart Lighting,80,8\n25-34,Smart Security Cameras,60,18\n35-44,Smart Thermostats,60,12\n35-44,Smart Lighting,68,10\n35-44,Smart Security Cameras,50,22\n45-54,Smart Thermostats,55,14\n45-54,Smart Lighting,62,12\n45-54,Smart Security Cameras,45,25\n55-64,Smart Thermostats,50,16\n55-64,Smart Lighting,58,15\n55-64,Smart Security Cameras,40,28\n65+,Smart Thermostats,40,20\n65+,Smart Lighting,50,18\n65+,Smart Security Cameras,35,30"},"code":{"kind":"string","value":"import numpy as np\nimport matplotlib.pyplot as plt\nfrom io import BytesIO\nfrom PIL import Image\n\ndef generate_plot(data):\n plt.style.use('default')\n \n age_groups = data['Age Group'].unique()\n device_categories = data['Device Category'].unique()\n x = np.arange(len(age_groups))\n width = 0.25\n \n fig, ax = plt.subplots(figsize=(12, 8))\n \n colors = ['#FF6B6B', '#4ECDC4', '#45B7D1']\n \n for i, device in enumerate(device_categories):\n popularity = data[data['Device Category'] == device]['Popularity (%)']\n error_rate = data[data['Device Category'] == device]['Installation Error Rate (%)']\n \n ax.bar(x + i*width, popularity, width, label=device, color=colors[i], alpha=0.8)\n ax.errorbar(x + i*width, popularity, yerr=error_rate, fmt='none', ecolor='black', capsize=5)\n \n ax.set_ylabel('Popularity (%)', fontsize=12, fontweight='bold')\n ax.set_xlabel('Age Group', fontsize=12, fontweight='bold')\n ax.set_title('Smart Home Device Popularity and Installation Error Rates', fontsize=16, fontweight='bold')\n ax.set_xticks(x + width)\n ax.set_xticklabels(age_groups, rotation=45, ha='right')\n ax.legend(title='Device Category', title_fontsize=12, fontsize=10, loc='upper right')\n \n ax.spines['top'].set_visible(False)\n ax.spines['right'].set_visible(False)\n \n ax.set_ylim(0, 100)\n ax.yaxis.grid(True, linestyle='--', alpha=0.7)\n \n plt.tight_layout()\n \n buffer = BytesIO()\n plt.savefig(buffer, format='png', dpi=300, bbox_inches='tight')\n buffer.seek(0)\n image = Image.open(buffer)\n \n return image"}}},{"rowIdx":15,"cells":{"id":{"kind":"string","value":"PlotlyChartPipeline_bar_claude_1-912"},"image":{"kind":"image","value":{"src":"https://datasets-server.huggingface.co/assets/allenai/CoSyn-400K/--/86e46e1fd5e754d056169f0fb38f06c6997ff7de/--/chart/train/15/image/image.jpg?Expires=1741922929&Signature=izVXeAncdFQPId7okSSHWmKTIkC53edOZyvXJLQ-R-TtieaDGwXjrYvtOZmTv6F~3IrzjLskgD~G-Y8dBy4ZSOb58yY7e7YKMzxNVZQmv-5kVg-LieIlPtm-xOu~-I8t~AnrfsNfX9CQKJkcHf2WxdnIYbAAIe4b1vTqQulWRrSf0IRgHxk0hdEkdVQafTvbTohBVRfFOvOJAcod6E-1fiM88AM4IFltYVLBrGJDZrXXeGqt7Mj-3ZwVRnevUy9h5KHdkurIxw7BcKw~4cpaS8y8jLmL887EPLVRaNAUIn1yD2FyKjjlsw0E02fiu56KwF~CIB9Kx50ENboUwEAOWQ__&Key-Pair-Id=K3EI6M078Z3AC3","height":1440,"width":2160},"inferredFromUrl":false},"qa_pairs":{"kind":"list like","value":{"question":["Which material is used most frequently overall by all the bird species?","Which bird species uses the highest percentage of spider silk in its nests?","Among the materials moss, grass, and twigs, which one does the American Robin use the most, and by what percentage?","How does the use of leaves by the Blue Tit compare to its use by the Tailorbird?","Which material has the least average use across all bird species?","How many bird species use mud in their nest-building?","What is the difference in twig usage between the Bald Eagle and Blue Tit?","Does any bird species not use grass in its nests? If yes, name the species.","What percentage of nest-building materials used by the Barn Swallow consists of mud and twigs combined?"],"explanation":["To determine the most frequently used material, look for the tallest bar in the figure regardless of the bird species. Each species' bar height represents the percentage use of a material. The material corresponding to the tallest bar is the one used most frequently.","Identify the bar labeled \"spider silk\" and note the height of this bar for each bird species. The species with the tallest bar for \"spider silk\" is the one that uses it the most.","Examine the bars for American Robin for moss, grass, and twigs. Compare their heights to identify which is tallest and note the corresponding percentage.","Locate the bar for leaves for both Blue Tit and Tailorbird. Compare the heights of these bars to determine which species uses more leaves.","Calculate the average usage of each material by summing the percentages for each material across all species and then dividing by the number of species (9). The material with the smallest average is the least used.","Count the number of species with a non-zero bar for mud. This is done by identifying the bars for mud and checking their height for each species.","Find the percentage of twigs used by the Bald Eagle and Blue Tit by looking at the height of the bars labeled for twigs. Subtract the smaller percentage from the larger one.","Check all bird species' bars for the material grass. Any species with a height of zero for grass indicates non-usage.","Locate the bars for mud and twigs for the Barn Swallow. Add their percentages to get the combined total."],"answer":["Twigs","Hummingbird","Twigs, 30%","The Tailorbird uses more leaves","Hair","3","60%","Tailorbird","80%"]},"string":"{\n \"question\": [\n \"Which material is used most frequently overall by all the bird species?\",\n \"Which bird species uses the highest percentage of spider silk in its nests?\",\n \"Among the materials moss, grass, and twigs, which one does the American Robin use the most, and by what percentage?\",\n \"How does the use of leaves by the Blue Tit compare to its use by the Tailorbird?\",\n \"Which material has the least average use across all bird species?\",\n \"How many bird species use mud in their nest-building?\",\n \"What is the difference in twig usage between the Bald Eagle and Blue Tit?\",\n \"Does any bird species not use grass in its nests? If yes, name the species.\",\n \"What percentage of nest-building materials used by the Barn Swallow consists of mud and twigs combined?\"\n ],\n \"explanation\": [\n \"To determine the most frequently used material, look for the tallest bar in the figure regardless of the bird species. Each species' bar height represents the percentage use of a material. The material corresponding to the tallest bar is the one used most frequently.\",\n \"Identify the bar labeled \\\"spider silk\\\" and note the height of this bar for each bird species. The species with the tallest bar for \\\"spider silk\\\" is the one that uses it the most.\",\n \"Examine the bars for American Robin for moss, grass, and twigs. Compare their heights to identify which is tallest and note the corresponding percentage.\",\n \"Locate the bar for leaves for both Blue Tit and Tailorbird. Compare the heights of these bars to determine which species uses more leaves.\",\n \"Calculate the average usage of each material by summing the percentages for each material across all species and then dividing by the number of species (9). The material with the smallest average is the least used.\",\n \"Count the number of species with a non-zero bar for mud. This is done by identifying the bars for mud and checking their height for each species.\",\n \"Find the percentage of twigs used by the Bald Eagle and Blue Tit by looking at the height of the bars labeled for twigs. Subtract the smaller percentage from the larger one.\",\n \"Check all bird species' bars for the material grass. Any species with a height of zero for grass indicates non-usage.\",\n \"Locate the bars for mud and twigs for the Barn Swallow. Add their percentages to get the combined total.\"\n ],\n \"answer\": [\n \"Twigs\",\n \"Hummingbird\",\n \"Twigs, 30%\",\n \"The Tailorbird uses more leaves\",\n \"Hair\",\n \"3\",\n \"60%\",\n \"Tailorbird\",\n \"80%\"\n ]\n}"},"metadata":{"kind":"string","value":"{\n \"figure_type\": \"bar chart\",\n \"persona\": \"A writer who shares captivating stories about extraordinary bird species and their survival strategies\",\n \"topic\": \"Comparison of nest-building materials used by various bird species\"\n}"},"data":{"kind":"string","value":"species,moss,twigs,grass,leaves,mud,spider_silk,lichen,feathers,hair\nAmerican Robin,10,30,15,5,35,0,0,5,0\nBald Eagle,5,80,10,5,0,0,0,0,0\nBarn Swallow,5,10,5,0,70,0,0,5,5\nBlue Tit,30,20,15,5,0,10,15,5,0\nHummingbird,5,5,5,5,0,70,10,0,0\nMagpie,5,70,10,5,10,0,0,0,0\nOvenbird,0,10,80,10,0,0,0,0,0\nSociable Weaver,0,60,30,0,0,0,0,5,5\nTailorbird,0,0,0,95,0,5,0,0,0\nWhite Stork,0,90,5,5,0,0,0,0,0"},"code":{"kind":"string","value":"import numpy as np\nimport plotly.graph_objects as go\nfrom io import BytesIO\nfrom PIL import Image\n\ndef generate_plot(df):\n materials = df.columns[1:].tolist()\n \n fig = go.Figure()\n \n for species in df['species']:\n fig.add_trace(go.Bar(\n name=species,\n x=materials,\n y=df[df['species'] == species].iloc[:, 1:].values.flatten(),\n marker_color=np.random.rand(3,),\n marker_line_color='rgb(8,48,107)',\n marker_line_width=1.5,\n opacity=0.8\n ))\n \n fig.update_layout(\n title={\n 'text': 'Nest-building Materials Used by Various Bird Species',\n 'font': {'size': 24, 'family': 'Arial', 'color': 'darkgreen'},\n 'x': 0.5,\n 'y': 0.95\n },\n xaxis_title={\n 'text': 'Materials',\n 'font': {'size': 18, 'family': 'Arial', 'color': 'darkblue'}\n },\n yaxis_title={\n 'text': 'Percentage Used',\n 'font': {'size': 18, 'family': 'Arial', 'color': 'darkblue'}\n },\n legend_title={\n 'text': 'Bird Species',\n 'font': {'size': 16, 'family': 'Arial', 'color': 'black'}\n },\n barmode='group',\n bargap=0.15,\n bargroupgap=0.1,\n plot_bgcolor='rgb(240,240,240)',\n font=dict(family=\"Arial\", size=12, color=\"black\"),\n legend=dict(\n x=1.05,\n y=1,\n bgcolor='rgba(255, 255, 255, 0.5)',\n bordercolor='rgba(0, 0, 0, 0.5)',\n borderwidth=1\n ),\n margin=dict(l=80, r=80, t=100, b=80),\n height=800,\n width=1200\n )\n \n fig.update_xaxes(tickangle=45)\n fig.update_yaxes(range=[0, 100])\n\n img_bytes = BytesIO()\n fig.write_image(img_bytes, format=\"png\", scale=2)\n img_bytes.seek(0)\n return Image.open(img_bytes)"}}},{"rowIdx":16,"cells":{"id":{"kind":"string","value":"MatplotlibChart_pie_claude_0-223"},"image":{"kind":"image","value":{"src":"https://datasets-server.huggingface.co/assets/allenai/CoSyn-400K/--/86e46e1fd5e754d056169f0fb38f06c6997ff7de/--/chart/train/16/image/image.jpg?Expires=1741922929&Signature=VIjgvClnimlhmWb2cwW~Wjbz-OylGX0J05ls~aE9ZSVt1RGepwia~vN2C2GN0KCyDbkrpNVgORCpliCaI39cqeUS-RVe74GjwNzZfPMj9dOnrCUx68Pd0rR9T6xjuzUPxBLl8feuZ8V51oP9UsYdIrKpHcrbIqsI3Lz-9YS1bJJHJRfYJRhKV9IcdZutbNnUA2Jh9S3Uopg5zSE8qzZQQsEbodmF55KE8FJ7Ip9auZyvhd7ou6WLQpIfcKpk5hW~RbSujXYLM1jMRkq7Sq5n2xgufUZFoJKUpe58uUQUpZQcmEg9Z8VTXG8MdzhtUkj~FeyiOa81QNIazvdzSgkmIQ__&Key-Pair-Id=K3EI6M078Z3AC3","height":1440,"width":1865},"inferredFromUrl":false},"qa_pairs":{"kind":"list like","value":{"question":["What proportion of genetic diseases are inherited through Autosomal dominant patterns?","Which inheritance pattern has the smallest representation in the pie chart?","How do the proportions of Autosomal recessive and Multifactorial diseases compare?","What are the combined percentages of X-linked recessive and Mitochondrial inheritance patterns?","How much larger is the Autosomal dominant proportion compared to the X-linked recessive proportion?","Which inheritance pattern is represented by a blue slice in the pie chart?","Add the proportions of all inheritance patterns except Autosomal dominant. What is the total?","Which inheritance pattern shares an equal proportion of 0.5% in genetic disease inheritance?","Is the percentage of Mitochondrial inheritance greater than that of Codominant inheritance?"],"explanation":["From the pie chart, find the sector labeled \"Autosomal dominant\" and read the percentage value presented.","Look for the sector with the smallest percentage value. The smallest slice is labeled \"Y-linked\" and \"X-linked dominant\", both contributing 0.5%.","Identify the percentages for both Autosomal recessive (25%) and Multifactorial (8%). Autosomal recessive has a higher proportion than Multifactorial.","Add the percentages of X-linked recessive (15%) and Mitochondrial (5%). 15% + 5% = 20%.","Subtract the percentage of X-linked recessive (15%) from Autosomal dominant (45%). 45% - 15% = 30%.","Identify the blue slice in the pie chart associated with a specific inheritance pattern label. The blue slice corresponds to \"Autosomal recessive\".","Add percentages of Autosomal recessive (25%), X-linked recessive (15%), Mitochondrial (5%), Multifactorial (8%), Y-linked (0.5%), Codominant (1%), and X-linked dominant (0.5%). 25% + 15% + 5% + 8% + 0.5% + 1% + 0.5% = 55%.","Identify the slices that each represent 0.5% and their associated labels. Both Y-linked and X-linked dominant have equal proportions of 0.5%.","Compare the percentage values of Mitochondrial (5%) and Codominant (1%). 5% is greater than 1%."],"answer":["45%","Y-linked and X-linked dominant","Autosomal recessive is greater","20%","30%","Autosomal recessive","55%","Y-linked and X-linked dominant","Yes"]},"string":"{\n \"question\": [\n \"What proportion of genetic diseases are inherited through Autosomal dominant patterns?\",\n \"Which inheritance pattern has the smallest representation in the pie chart?\",\n \"How do the proportions of Autosomal recessive and Multifactorial diseases compare?\",\n \"What are the combined percentages of X-linked recessive and Mitochondrial inheritance patterns?\",\n \"How much larger is the Autosomal dominant proportion compared to the X-linked recessive proportion?\",\n \"Which inheritance pattern is represented by a blue slice in the pie chart?\",\n \"Add the proportions of all inheritance patterns except Autosomal dominant. What is the total?\",\n \"Which inheritance pattern shares an equal proportion of 0.5% in genetic disease inheritance?\",\n \"Is the percentage of Mitochondrial inheritance greater than that of Codominant inheritance?\"\n ],\n \"explanation\": [\n \"From the pie chart, find the sector labeled \\\"Autosomal dominant\\\" and read the percentage value presented.\",\n \"Look for the sector with the smallest percentage value. The smallest slice is labeled \\\"Y-linked\\\" and \\\"X-linked dominant\\\", both contributing 0.5%.\",\n \"Identify the percentages for both Autosomal recessive (25%) and Multifactorial (8%). Autosomal recessive has a higher proportion than Multifactorial.\",\n \"Add the percentages of X-linked recessive (15%) and Mitochondrial (5%). 15% + 5% = 20%.\",\n \"Subtract the percentage of X-linked recessive (15%) from Autosomal dominant (45%). 45% - 15% = 30%.\",\n \"Identify the blue slice in the pie chart associated with a specific inheritance pattern label. The blue slice corresponds to \\\"Autosomal recessive\\\".\",\n \"Add percentages of Autosomal recessive (25%), X-linked recessive (15%), Mitochondrial (5%), Multifactorial (8%), Y-linked (0.5%), Codominant (1%), and X-linked dominant (0.5%). 25% + 15% + 5% + 8% + 0.5% + 1% + 0.5% = 55%.\",\n \"Identify the slices that each represent 0.5% and their associated labels. Both Y-linked and X-linked dominant have equal proportions of 0.5%.\",\n \"Compare the percentage values of Mitochondrial (5%) and Codominant (1%). 5% is greater than 1%.\"\n ],\n \"answer\": [\n \"45%\",\n \"Y-linked and X-linked dominant\",\n \"Autosomal recessive is greater\",\n \"20%\",\n \"30%\",\n \"Autosomal recessive\",\n \"55%\",\n \"Y-linked and X-linked dominant\",\n \"Yes\"\n ]\n}"},"metadata":{"kind":"string","value":"{\n \"figure_type\": \"pie chart\",\n \"persona\": \"A biology professor teaching a course on genetics and using real-life case studies, including inherited eye conditions, to illustrate concepts\",\n \"topic\": \"Proportion of different inheritance patterns in human genetic diseases\"\n}"},"data":{"kind":"string","value":"inheritance_pattern,percentage\nAutosomal dominant,45\nAutosomal recessive,25\nX-linked recessive,15\nMitochondrial,5\nMultifactorial,8\nY-linked,0.5\nCodominant,1\nX-linked dominant,0.5"},"code":{"kind":"string","value":"import numpy as np\nimport matplotlib.pyplot as plt\nfrom io import BytesIO\nfrom PIL import Image\n\ndef generate_plot(data):\n plt.style.use('default')\n \n # Prepare data\n labels = data['inheritance_pattern']\n sizes = data['percentage']\n \n # Define colors\n colors = ['#FF9999', '#66B2FF', '#99FF99', '#FFCC99', '#FF99CC', '#FFD700', '#00CED1', '#9370DB']\n \n # Create figure and axis\n fig, ax = plt.subplots(figsize=(10, 8))\n \n # Create pie chart\n wedges, texts, autotexts = ax.pie(sizes, labels=labels, colors=colors, \n autopct='%1.1f%%', startangle=90, \n pctdistance=0.85, labeldistance=1.05)\n \n # Customize wedges\n for w in wedges:\n w.set_edgecolor('white')\n w.set_linewidth(2)\n \n # Customize text\n for text in texts:\n text.set_fontsize(10)\n text.set_fontweight('bold')\n \n for autotext in autotexts:\n autotext.set_fontsize(9)\n autotext.set_fontweight('bold')\n \n # Add title\n ax.set_title('Proportion of Different Inheritance Patterns\\nin Human Genetic Diseases', \n fontsize=16, fontweight='bold', pad=20)\n \n # Add legend\n ax.legend(wedges, labels, title=\"Inheritance Patterns\", \n loc=\"center left\", bbox_to_anchor=(1, 0, 0.5, 1))\n \n # Adjust layout\n plt.tight_layout()\n \n # Save plot to BytesIO\n buf = BytesIO()\n plt.savefig(buf, format='png', dpi=300, bbox_inches='tight')\n buf.seek(0)\n \n # Convert to PIL Image\n img = Image.open(buf)\n \n return img"}}},{"rowIdx":17,"cells":{"id":{"kind":"string","value":"PlotlyChartPipeline_gauge_1-443"},"image":{"kind":"image","value":{"src":"https://datasets-server.huggingface.co/assets/allenai/CoSyn-400K/--/86e46e1fd5e754d056169f0fb38f06c6997ff7de/--/chart/train/17/image/image.jpg?Expires=1741922929&Signature=xH45~OpWA0Hh5nakcPkCf288~6jkwcpS3VhM-gsOCv9MVqJupBfKwDDQRuPSwy-Hf4YbMexwuk9FcqLZEts6t8TLBeVpTPhEmZFmtEEeAtPm1mckX4Xgaq9SV5vDyXKgWhDn5Z2gZLTpo4zp0y-uwh7RvIskgvoiAuztbOpjdhSOZEg83KTEbmWZBCPms1Jp3KVsrsqnwsSH3ka2gfKrqiQkwKf5z5ezIUl1-LNk1yvihxJ7mZl29OOmbjaHlSztjSl1zYQFpHREXWL1WKkO925Dh7h5JRR1ZE1VeaDiOFcHmInZskfM5kFdkbv94fCnL03fWNRcLWop3KMkbaAUfg__&Key-Pair-Id=K3EI6M078Z3AC3","height":1440,"width":2160},"inferredFromUrl":false},"qa_pairs":{"kind":"list like","value":{"question":["Which solar panel type has the highest efficiency percentage?","What is the range of efficiency percentages across all solar panel types?","Compare the efficiency of Monocrystalline and Polycrystalline solar panels. Which one is higher, and by how much?","What is the average efficiency percentage of all the types of solar panels?","How much more efficient is Bifacial compared to Thin-Film solar panels?","Which solar panel type has an efficiency closest to the average efficiency?","What is the total combined efficiency percentage of Monocrystalline and Perovskite solar panels?","What is the percentage difference between the panel types with the highest and lowest efficiency?","Out of the given solar panel types, which ones have an efficiency below 20%?"],"explanation":["The gauge chart shows five types of solar panels with efficiency percentages. The highest value is indicated by the Perovskite type, which shows 28%.","The lowest efficiency percentage is 12% (Thin-Film) and the highest is 28% (Perovskite). Therefore, the range is 28% - 12% = 16%.","Monocrystalline has an efficiency of 22%, and Polycrystalline has 18%. The difference is 22% - 18% = 4%.","Add all the efficiency percentages: 22% + 18% + 12% + 25% + 28% = 105%. The average is 105% / 5 = 21%.","Bifacial has an efficiency of 25%, and Thin-Film has 12%. The difference is 25% - 12% = 13%.","The average efficiency percentage is 21%. The closest value to 21% is Monocrystalline with 22%.","Monocrystalline's efficiency is 22%, and Perovskite's is 28%. The combined efficiency is 22% + 28% = 50%.","The highest efficiency is 28% (Perovskite), and the lowest is 12% (Thin-Film). The percentage difference is (28% - 12%) / 12% * 100% = 133.33%.","The gauge chart shows Polycrystalline with 18% and Thin-Film with 12% as the types with efficiency below 20%."],"answer":["Perovskite","16%","Monocrystalline by 4%","21%","13%","Monocrystalline","50%","133.33%","Polycrystalline and Thin-Film"]},"string":"{\n \"question\": [\n \"Which solar panel type has the highest efficiency percentage?\",\n \"What is the range of efficiency percentages across all solar panel types?\",\n \"Compare the efficiency of Monocrystalline and Polycrystalline solar panels. Which one is higher, and by how much?\",\n \"What is the average efficiency percentage of all the types of solar panels?\",\n \"How much more efficient is Bifacial compared to Thin-Film solar panels?\",\n \"Which solar panel type has an efficiency closest to the average efficiency?\",\n \"What is the total combined efficiency percentage of Monocrystalline and Perovskite solar panels?\",\n \"What is the percentage difference between the panel types with the highest and lowest efficiency?\",\n \"Out of the given solar panel types, which ones have an efficiency below 20%?\"\n ],\n \"explanation\": [\n \"The gauge chart shows five types of solar panels with efficiency percentages. The highest value is indicated by the Perovskite type, which shows 28%.\",\n \"The lowest efficiency percentage is 12% (Thin-Film) and the highest is 28% (Perovskite). Therefore, the range is 28% - 12% = 16%.\",\n \"Monocrystalline has an efficiency of 22%, and Polycrystalline has 18%. The difference is 22% - 18% = 4%.\",\n \"Add all the efficiency percentages: 22% + 18% + 12% + 25% + 28% = 105%. The average is 105% / 5 = 21%.\",\n \"Bifacial has an efficiency of 25%, and Thin-Film has 12%. The difference is 25% - 12% = 13%.\",\n \"The average efficiency percentage is 21%. The closest value to 21% is Monocrystalline with 22%.\",\n \"Monocrystalline's efficiency is 22%, and Perovskite's is 28%. The combined efficiency is 22% + 28% = 50%.\",\n \"The highest efficiency is 28% (Perovskite), and the lowest is 12% (Thin-Film). The percentage difference is (28% - 12%) / 12% * 100% = 133.33%.\",\n \"The gauge chart shows Polycrystalline with 18% and Thin-Film with 12% as the types with efficiency below 20%.\"\n ],\n \"answer\": [\n \"Perovskite\",\n \"16%\",\n \"Monocrystalline by 4%\",\n \"21%\",\n \"13%\",\n \"Monocrystalline\",\n \"50%\",\n \"133.33%\",\n \"Polycrystalline and Thin-Film\"\n ]\n}"},"metadata":{"kind":"string","value":"{\n \"figure_type\": \"Gauge Chart (only display one number on each gauge without text overlapping)\",\n \"persona\": \"A former meteorologist who admires the beauty and power of natural phenomena, like wind turbines\",\n \"topic\": \"Solar panel energy conversion rate\"\n}"},"data":{"kind":"string","value":"Solar_Panel_Type,Efficiency_Percentage\nMonocrystalline,22\nPolycrystalline,18\nThin-Film,12\nBifacial,25\nPerovskite,28"},"code":{"kind":"string","value":"import plotly.graph_objects as go\nfrom plotly.subplots import make_subplots\nimport pandas as pd\nimport numpy as np\nfrom io import BytesIO\nfrom PIL import Image\n\ndef generate_plot(df):\n # Create subplots\n fig = make_subplots(\n rows=2, cols=3,\n specs=[[{'type': 'indicator'}, {'type': 'indicator'}, {'type': 'indicator'}],\n [{'type': 'indicator'}, {'type': 'indicator'}, None]],\n subplot_titles=df['Solar_Panel_Type']\n )\n\n # Custom color palette inspired by solar panels and nature\n colors = ['#4a7fb5', '#6fb98f', '#ffd700', '#ff7f50', '#2e8b57']\n\n # Add gauges for each solar panel type\n for i, (_, row) in enumerate(df.iterrows()):\n fig.add_trace(go.Indicator(\n mode=\"gauge+number\",\n value=row['Efficiency_Percentage'],\n domain={'row': i // 3, 'column': i % 3},\n title={'text': row['Solar_Panel_Type'], 'font': {'size': 24, 'color': colors[i]}},\n gauge={\n 'axis': {'range': [None, 30], 'tickwidth': 1, 'tickcolor': colors[i]},\n 'bar': {'color': colors[i]},\n 'bgcolor': \"white\",\n 'borderwidth': 2,\n 'bordercolor': colors[i],\n 'steps': [\n {'range': [0, row['Efficiency_Percentage']], 'color': f'rgba{tuple(int(colors[i].lstrip(\"#\")[i:i+2], 16) for i in (0, 2, 4)) + (0.2,)}'}\n ],\n }\n ), row=i // 3 + 1, col=i % 3 + 1)\n\n # Update layout\n fig.update_layout(\n title={\n 'text': \"Solar Panel Efficiency Comparison\",\n 'y': 0.95,\n 'x': 0.5,\n 'xanchor': 'center',\n 'yanchor': 'top',\n 'font': {'size': 30, 'color': '#1a3a5a'}\n },\n font={'family': \"Arial\", 'size': 12},\n paper_bgcolor=\"rgba(240,240,240,0.95)\",\n plot_bgcolor=\"rgba(240,240,240,0.95)\",\n height=800,\n width=1200,\n margin=dict(l=50, r=50, t=100, b=50)\n )\n\n # Save plot to BytesIO object\n img_bytes = BytesIO()\n fig.write_image(img_bytes, format=\"png\", scale=2)\n img_bytes.seek(0)\n \n # Return plot as PIL Image object\n return Image.open(img_bytes)"}}},{"rowIdx":18,"cells":{"id":{"kind":"string","value":"MatplotlibChart_specific_0-66"},"image":{"kind":"image","value":{"src":"https://datasets-server.huggingface.co/assets/allenai/CoSyn-400K/--/86e46e1fd5e754d056169f0fb38f06c6997ff7de/--/chart/train/18/image/image.jpg?Expires=1741922929&Signature=r8eb9jcKy0o6tOQ1BoxqbHuxgPVsluzzJIcsWNjRv2Ms5XouoFhbDskpMxpwMvopFV~CnVP4Fq~2k-479VjOivnZuUplCyZEwzPZnNwmuW5F4bmvFMRkjjqq40TuLZi0nhuX9R7CgSFl-NfG1ACd5ZFG1RMZ1dXXLy3Ng3HDlz4Nr7TaGjZIhzNYpEoThFpxE19kaHuJaZeTu5setwpx~kP2P3vSW6YA3Meb9uax7B4GXDAAYsItmiVswfuaPEAYOSpvWakK8mDsPRhAIyqxB7HQ~Tn5aTh4CsKOqMIoWNONI8p2M8n2fDl2xDiQFm-1Sd-Zxy5i6lWd~UwTwoFkgw__&Key-Pair-Id=K3EI6M078Z3AC3","height":590,"width":1207},"inferredFromUrl":false},"qa_pairs":{"kind":"list like","value":{"question":["Which TV series had the earliest finale date?","Which TV series had finales in the same year and what were the pivotal moments?","How many TV series finales occurred after 2018?","Which series' finale involved a major character's demise?","Which TV series finale has the longest title for its pivotal moment?"],"explanation":["Look at the dates alongside the TV series; find the earliest one. That's \"Breaking Bad\" on 2013-09-29.","Identify series that share the same year value. For 2020, it's \"BoJack Horseman\" and \"The Good Place\". Their pivotal moments are \"BoJack and Diane's final conversation\" and \"The characters' final decisions about their afterlife\" respectively.","Count the number of series finales with a year greater than 2018. There are four: \"Game of Thrones\", \"Shameless\", and \"Ozark\".","Look for moments that signify a character's death. \"Breaking Bad\" (Walter White's demise), \"The Vampire Diaries\" (Stefan's sacrifice), and \"Shameless\" (Frank Gallagher's final moments) all fit this criterion.","Compare the lengths of each pivotal moment description. \"Revelation of the mother's fate and Ted's relationship with Robin\" from \"How I Met Your Mother\" is the longest."],"answer":["Breaking Bad","BoJack Horseman: BoJack and Diane's final conversation; The Good Place: The characters' final decisions about their afterlife","Four","Breaking Bad, The Vampire Diaries, Shameless","How I Met Your Mother"]},"string":"{\n \"question\": [\n \"Which TV series had the earliest finale date?\",\n \"Which TV series had finales in the same year and what were the pivotal moments?\",\n \"How many TV series finales occurred after 2018?\",\n \"Which series' finale involved a major character's demise?\",\n \"Which TV series finale has the longest title for its pivotal moment?\"\n ],\n \"explanation\": [\n \"Look at the dates alongside the TV series; find the earliest one. That's \\\"Breaking Bad\\\" on 2013-09-29.\",\n \"Identify series that share the same year value. For 2020, it's \\\"BoJack Horseman\\\" and \\\"The Good Place\\\". Their pivotal moments are \\\"BoJack and Diane's final conversation\\\" and \\\"The characters' final decisions about their afterlife\\\" respectively.\",\n \"Count the number of series finales with a year greater than 2018. There are four: \\\"Game of Thrones\\\", \\\"Shameless\\\", and \\\"Ozark\\\".\",\n \"Look for moments that signify a character's death. \\\"Breaking Bad\\\" (Walter White's demise), \\\"The Vampire Diaries\\\" (Stefan's sacrifice), and \\\"Shameless\\\" (Frank Gallagher's final moments) all fit this criterion.\",\n \"Compare the lengths of each pivotal moment description. \\\"Revelation of the mother's fate and Ted's relationship with Robin\\\" from \\\"How I Met Your Mother\\\" is the longest.\"\n ],\n \"answer\": [\n \"Breaking Bad\",\n \"BoJack Horseman: BoJack and Diane's final conversation; The Good Place: The characters' final decisions about their afterlife\",\n \"Four\",\n \"Breaking Bad, The Vampire Diaries, Shameless\",\n \"How I Met Your Mother\"\n ]\n}"},"metadata":{"kind":"string","value":"{\n \"figure_type\": \"event plot\",\n \"persona\": \"a cable-TV-loving, stay-at-home mom in her late forties\",\n \"topic\": \"Pivotal moments in popular TV series finales over the past decade by year\"\n}"},"data":{"kind":"string","value":"Year,TV Series,Finale Date,Pivotal Moment\n2013,Breaking Bad,2013-09-29,Walter White's final showdown and demise\n2014,How I Met Your Mother,2014-03-31,Revelation of the mother's fate and Ted's relationship with Robin\n2015,Mad Men,2015-05-17,Don Draper's realization and the iconic Coca-Cola commercial\n2016,The Good Wife,2016-05-08,Alicia Florrick's betrayal and slap\n2017,The Vampire Diaries,2017-03-10,Stefan's sacrifice to save Mystic Falls\n2018,The Americans,2018-05-30,The Jennings' escape and FBI confrontation\n2019,Game of Thrones,2019-05-19,Bran Stark being named king and Jon Snow's exile\n2020,BoJack Horseman,2020-01-31,BoJack and Diane's final conversation\n2020,The Good Place,2020-01-30,The characters' final decisions about their afterlife\n2021,Shameless,2021-04-11,Frank Gallagher's final moments and the family gathering\n2022,Ozark,2022-04-29,The Byrde family's moral reckoning and Ruth Langmore's fate"},"code":{"kind":"string","value":"import numpy as np\nimport matplotlib.pyplot as plt\nimport pandas as pd\nfrom io import BytesIO\nfrom PIL import Image\n\ndef generate_plot(df):\n # Setting style\n plt.style.use('ggplot')\n \n # Extracting data\n years = df['Year']\n pivotal_moments = df['Pivotal Moment']\n tv_series = df['TV Series']\n \n # Figure and axis\n fig, ax = plt.subplots(figsize=(12, 6))\n \n # Plotting events\n y_positions = np.arange(len(years))\n ax.scatter(years, y_positions, color='purple', marker='o', s=100, label='Finale Events')\n \n # Annotating pivotal moments\n for i in range(len(years)):\n ax.text(years[i], y_positions[i], f\"{tv_series[i]}: {pivotal_moments[i]}\", \n verticalalignment='center', fontsize=10, color='black')\n \n # Customizing the plot\n ax.set_yticks(y_positions)\n ax.set_yticklabels(tv_series)\n ax.set_xlabel('Year')\n ax.set_title('Pivotal Moments in Popular TV Series Finales (2013-2022)')\n \n plt.tight_layout()\n \n # Saving to BytesIO\n buf = BytesIO()\n plt.savefig(buf, format='png', bbox_inches='tight')\n \n # Returning as PIL image\n buf.seek(0)\n img = Image.open(buf)\n \n return img\n\nimport pandas as pd\n\n# Sample DataFrame (from provided data)\ndata = {\n 'Year': [2013, 2014, 2015, 2016, 2017, 2018, 2019, 2020, 2020, 2021, 2022],\n 'TV Series': ['Breaking Bad', 'How I Met Your Mother', 'Mad Men', 'The Good Wife', 'The Vampire Diaries',\n 'The Americans', 'Game of Thrones', 'BoJack Horseman', 'The Good Place', 'Shameless', 'Ozark'],\n 'Finale Date': ['2013-09-29', '2014-03-31', '2015-05-17', '2016-05-08', '2017-03-10', \n '2018-05-30', '2019-05-19', '2020-01-31', '2020-01-30', '2021-04-11', '2022-04-29'],\n 'Pivotal Moment': [\"Walter White's final showdown and demise\",\n \"Revelation of the mother's fate and Ted's relationship with Robin\",\n \"Don Draper's realization and the iconic Coca-Cola commercial\",\n \"Alicia Florrick's betrayal and slap\",\n \"Stefan's sacrifice to save Mystic Falls\",\n \"The Jennings' escape and FBI confrontation\",\n \"Bran Stark being named king and Jon Snow's exile\",\n \"BoJack and Diane's final conversation\",\n \"The characters' final decisions about their afterlife\",\n \"Frank Gallagher's final moments and the family gathering\",\n \"The Byrde family's moral reckoning and Ruth Langmore's fate\"]\n}\n\ndf = pd.DataFrame(data)"}}},{"rowIdx":19,"cells":{"id":{"kind":"string","value":"MatplotlibChart_specific_1-556"},"image":{"kind":"image","value":{"src":"https://datasets-server.huggingface.co/assets/allenai/CoSyn-400K/--/86e46e1fd5e754d056169f0fb38f06c6997ff7de/--/chart/train/19/image/image.jpg?Expires=1741922929&Signature=j1Hz3g2XNNP2ZNNgwdqMtfjH46oOK9Ww5fn8NfLUoH4Ufa142yI3GFdkOniRU9yIJbyEVANZB1P6rNfe94fOr~PrxiqnJHRiC-QOKtE3FEt9y62r2ghasXBiurPwk2I1v6x5fPCpYtFm18Lyi3FTLy~uGqCF2rtNq0ao7dh~H~yjFLaJyHUicuA4QgTt7w~zfgekCtK8noBJt0ofWeAfzW0va7idVEPCOCZeQpGsy8rX7oq6QBjrml~iXIPxJ9fM~D87FTuvyK9gWkW-04oWl5FX1IXEpRjFQPdFxxPQegXve5u8LsX08w4Hg1bVtK1UrHCoLlJ95ObUkZcT5p6Cig__&Key-Pair-Id=K3EI6M078Z3AC3","height":590,"width":989},"inferredFromUrl":false},"qa_pairs":{"kind":"list like","value":{"question":["What's the pattern in the trend of nanomaterial adoption in textile manufacturing from 2013 to 2023?","How much did the adoption percentage increase from 2019 to 2020?","Which year had the highest nanomaterial adoption percentage?","By what factor did the adoption percentage increase from 2013 to 2023?","Was the increase in adoption percentage from 2017 to 2018 greater than from 2018 to 2019?","Which year saw the largest single-year increase in adoption percentage?","What's the average nanomaterial adoption percentage over the last decade?","Did the adoption percentage increase by more than 10% between any consecutive years?","What was the adoption percentage halfway between the years 2015 and 2018?"],"explanation":["Observing the curve in the figure, the adoption percentage consistently increases every year, indicating a steadily rising trend.","Subtract the adoption percentage in 2019 (27%) from the adoption percentage in 2020 (34%). The difference is 34 - 27 = 7.","From the figure, the adoption percentage is plotted against years, and 2023 is the highest point at 59%.","Divide the adoption percentage in 2023 (59%) by that in 2013 (5%). The factor is 59 / 5 = 11.8.","Calculate the increase from 2017 to 2018 (21% - 16% = 5%) and from 2018 to 2019 (27% - 21% = 6%). Comparing these, 6% is greater than 5%.","Comparing the yearly increases: \n2013 to 2014: 2%, \n2014 to 2015: 2%, \n2015 to 2016: 3%, \n2016 to 2017: 4%, \n2017 to 2018: 5%, \n2018 to 2019: 6%, \n2019 to 2020: 7%, \n2020 to 2021: 8%, \n2021 to 2022: 8%, \n2022 to 2023: 9%. The largest increase is 9% from 2022 to 2023.","Sum the adoption percentages from 2013 to 2023 and divide by the number of years. Total sum = 5 + 7 + 9 + 12 + 16 + 21 + 27 + 34 + 42 + 50 + 59 = 282. Average = 282/11 = approximately 25.6%.","Check each year's increase:\n2013 to 2014: 2%,\n2014 to 2015: 2%,\n2015 to 2016: 3%,\n2016 to 2017: 4%,\n2017 to 2018: 5%,\n2018 to 2019: 6%,\n2019 to 2020: 7%,\n2020 to 2021: 8%,\n2021 to 2022: 8%,\n2022 to 2023: 9%. No increase is greater than 10%.","For halfway between years, calculate the midpoint. The midpoint year between 2015 and 2018 is 2016.5 (which rounds to 2017). The adoption percentage in 2017 is 16%."],"answer":["The adoption percentage increases every year","7%","2023","11.8","No","2022 to 2023","Approximately 25.6%","No","16%"]},"string":"{\n \"question\": [\n \"What's the pattern in the trend of nanomaterial adoption in textile manufacturing from 2013 to 2023?\",\n \"How much did the adoption percentage increase from 2019 to 2020?\",\n \"Which year had the highest nanomaterial adoption percentage?\",\n \"By what factor did the adoption percentage increase from 2013 to 2023?\",\n \"Was the increase in adoption percentage from 2017 to 2018 greater than from 2018 to 2019?\",\n \"Which year saw the largest single-year increase in adoption percentage?\",\n \"What's the average nanomaterial adoption percentage over the last decade?\",\n \"Did the adoption percentage increase by more than 10% between any consecutive years?\",\n \"What was the adoption percentage halfway between the years 2015 and 2018?\"\n ],\n \"explanation\": [\n \"Observing the curve in the figure, the adoption percentage consistently increases every year, indicating a steadily rising trend.\",\n \"Subtract the adoption percentage in 2019 (27%) from the adoption percentage in 2020 (34%). The difference is 34 - 27 = 7.\",\n \"From the figure, the adoption percentage is plotted against years, and 2023 is the highest point at 59%.\",\n \"Divide the adoption percentage in 2023 (59%) by that in 2013 (5%). The factor is 59 / 5 = 11.8.\",\n \"Calculate the increase from 2017 to 2018 (21% - 16% = 5%) and from 2018 to 2019 (27% - 21% = 6%). Comparing these, 6% is greater than 5%.\",\n \"Comparing the yearly increases: \\n2013 to 2014: 2%, \\n2014 to 2015: 2%, \\n2015 to 2016: 3%, \\n2016 to 2017: 4%, \\n2017 to 2018: 5%, \\n2018 to 2019: 6%, \\n2019 to 2020: 7%, \\n2020 to 2021: 8%, \\n2021 to 2022: 8%, \\n2022 to 2023: 9%. The largest increase is 9% from 2022 to 2023.\",\n \"Sum the adoption percentages from 2013 to 2023 and divide by the number of years. Total sum = 5 + 7 + 9 + 12 + 16 + 21 + 27 + 34 + 42 + 50 + 59 = 282. Average = 282/11 = approximately 25.6%.\",\n \"Check each year's increase:\\n2013 to 2014: 2%,\\n2014 to 2015: 2%,\\n2015 to 2016: 3%,\\n2016 to 2017: 4%,\\n2017 to 2018: 5%,\\n2018 to 2019: 6%,\\n2019 to 2020: 7%,\\n2020 to 2021: 8%,\\n2021 to 2022: 8%,\\n2022 to 2023: 9%. No increase is greater than 10%.\",\n \"For halfway between years, calculate the midpoint. The midpoint year between 2015 and 2018 is 2016.5 (which rounds to 2017). The adoption percentage in 2017 is 16%.\"\n ],\n \"answer\": [\n \"The adoption percentage increases every year\",\n \"7%\",\n \"2023\",\n \"11.8\",\n \"No\",\n \"2022 to 2023\",\n \"Approximately 25.6%\",\n \"No\",\n \"16%\"\n ]\n}"},"metadata":{"kind":"string","value":"{\n \"figure_type\": \"plot of math functions\",\n \"persona\": \"A textile manufacturer offering a range of nanomaterial-treated fabrics\",\n \"topic\": \"Trend Analysis of Nanomaterial Adoption in Textile Manufacturing Over the Last Decade\"\n}"},"data":{"kind":"string","value":"Year,Nanomaterial_Adoption_Percentage\n2013,5\n2014,7\n2015,9\n2016,12\n2017,16\n2018,21\n2019,27\n2020,34\n2021,42\n2022,50\n2023,59"},"code":{"kind":"string","value":"import numpy as np\nimport matplotlib.pyplot as plt\nimport pandas as pd\nfrom io import BytesIO\nfrom PIL import Image\n\ndef generate_plot(data):\n # Applying style\n plt.style.use('seaborn-v0_8-white')\n\n # Extracting data\n years = data['Year']\n adoption_percentages = data['Nanomaterial_Adoption_Percentage']\n \n # Creating figure and axis objects\n fig, ax = plt.subplots(figsize=(10, 6))\n \n # Plotting data\n ax.plot(years, adoption_percentages, marker='o', color='teal', markersize=8, linewidth=2, label='Adoption Percentage')\n \n # Customizing plot\n ax.set_title('Nanomaterial Adoption in Textile Manufacturing (2013-2023)', fontsize=16, fontweight='bold', color='darkblue')\n ax.set_xlabel('Year', fontsize=14, fontweight='bold')\n ax.set_ylabel('Adoption Percentage (%)', fontsize=14, fontweight='bold')\n ax.set_xticks(np.arange(2013, 2024, 1))\n ax.set_yticks(np.arange(0, 70, 10))\n \n ax.grid(True, linestyle='--', alpha=0.7)\n ax.legend(loc='upper left', fontsize=12)\n ax.tick_params(axis='both', which='major', labelsize=12)\n \n # Adjust layout and save plot to BytesIO object\n plt.tight_layout()\n plot_stream = BytesIO()\n plt.savefig(plot_stream, format='png', bbox_inches='tight', dpi=100)\n \n plot_stream.seek(0)\n image = Image.open(plot_stream)\n \n return image"}}},{"rowIdx":20,"cells":{"id":{"kind":"string","value":"PlotlyChartPipeline_funnel_1-118"},"image":{"kind":"image","value":{"src":"https://datasets-server.huggingface.co/assets/allenai/CoSyn-400K/--/86e46e1fd5e754d056169f0fb38f06c6997ff7de/--/chart/train/20/image/image.jpg?Expires=1741922929&Signature=YbidU64QqzDMT6VSLwjgffTzTBcI52nCvEC8hwc9G3jSLt6r5cWm-jy3asukeXCMwBur4U934260iV5zjDma42Q71bCZps-4F7ZJyAOChtH63mT2f3wyum6SS9scJj~EDa0YZB1lsJ9kZroLDRMN7RWQktjUOapf--9EilLijtrUxh-~a0aeQWhHrvp~pS2A4ab91rcNofAFo9hdQQEi-Ds~7Sq~Q0xl8w1z7w-kV3IzBK~bKmgA4gF6NhJ2tbMff13PiDLS8qQQw68gdzeAt6RK7NV8fcVVEwjCNF2P5NX4Qy15VZu9t05yNzbgZwfftDNGWW6QoLiQtQQECYJ9wg__&Key-Pair-Id=K3EI6M078Z3AC3","height":1400,"width":1800},"inferredFromUrl":false},"qa_pairs":{"kind":"list like","value":{"question":["What is the title of the funnel chart?","Which country has the highest percentage of protected land?","What is the percentage of protected land in France?","How many countries have more than 20% of their land protected?","Which country has slightly more protected land percentage, China or Mexico?","What is the total percentage of protected land for Brazil and France combined?","Which country has the lowest percentage of protected land?","Compare the protected land percentages of Canada and Russia. Which is higher and by how much?","What's the average protected land percentage among the ten countries?","What is the median percentage of protected land among these countries?"],"explanation":["The title is located at the top center of the funnel chart, written in larger, bold, and colorful text. It reads \"Protected Land Percentage by Country.\"","The country with the highest percentage of protected land is at the widest portion at the top of the funnel. In this case, it is Germany with 37.8%.","France is located second from the top in the funnel chart. The percentage next to France is 28.1%.","By looking at the percentages listed next to each country in the funnel chart, count those countries with percentages above 20%. They are Germany, Brazil, and France, which totals three countries.","Locate China and Mexico in the funnel chart and compare their percentages. China has 15.6% while Mexico has 14.2%, so China has a slightly higher percentage.","Brazil has a protected land percentage of 29.4% and France has 28.1%. Adding them together results in 29.4 + 28.1 = 57.5%.","The country at the narrowest portion of the funnel, at the bottom, represents the lowest percentage. It is India with 5.9%.","Canada has 11.8% protected land and Russia has 9.7%. Canada has a higher percentage. The difference is 11.8 - 9.7 = 2.1%.","Sum the percentages of all ten countries and then divide by the number of countries (10). The total is 29.4 + 13.9 + 19.7 + 15.6 + 11.8 + 9.7 + 5.9 + 14.2 + 37.8 + 28.1 = 186.1%. The average is 186.1 / 10 = 18.61%.","To find the median, list the percentages in order and identify the middle number. The ordered percentages are 5.9, 9.7, 11.8, 13.9, 14.2, 15.6, 19.7, 28.1, 29.4, 37.8. The middle values are 14.2 and 15.6. The median is the average of these two: (14.2 + 15.6) / 2 = 14.9%."],"answer":["Protected Land Percentage by Country","Germany","28.1%","3","China","57.5%","India","Canada, 2.1%","18.61%","14.9%"]},"string":"{\n \"question\": [\n \"What is the title of the funnel chart?\",\n \"Which country has the highest percentage of protected land?\",\n \"What is the percentage of protected land in France?\",\n \"How many countries have more than 20% of their land protected?\",\n \"Which country has slightly more protected land percentage, China or Mexico?\",\n \"What is the total percentage of protected land for Brazil and France combined?\",\n \"Which country has the lowest percentage of protected land?\",\n \"Compare the protected land percentages of Canada and Russia. Which is higher and by how much?\",\n \"What's the average protected land percentage among the ten countries?\",\n \"What is the median percentage of protected land among these countries?\"\n ],\n \"explanation\": [\n \"The title is located at the top center of the funnel chart, written in larger, bold, and colorful text. It reads \\\"Protected Land Percentage by Country.\\\"\",\n \"The country with the highest percentage of protected land is at the widest portion at the top of the funnel. In this case, it is Germany with 37.8%.\",\n \"France is located second from the top in the funnel chart. The percentage next to France is 28.1%.\",\n \"By looking at the percentages listed next to each country in the funnel chart, count those countries with percentages above 20%. They are Germany, Brazil, and France, which totals three countries.\",\n \"Locate China and Mexico in the funnel chart and compare their percentages. China has 15.6% while Mexico has 14.2%, so China has a slightly higher percentage.\",\n \"Brazil has a protected land percentage of 29.4% and France has 28.1%. Adding them together results in 29.4 + 28.1 = 57.5%.\",\n \"The country at the narrowest portion of the funnel, at the bottom, represents the lowest percentage. It is India with 5.9%.\",\n \"Canada has 11.8% protected land and Russia has 9.7%. Canada has a higher percentage. The difference is 11.8 - 9.7 = 2.1%.\",\n \"Sum the percentages of all ten countries and then divide by the number of countries (10). The total is 29.4 + 13.9 + 19.7 + 15.6 + 11.8 + 9.7 + 5.9 + 14.2 + 37.8 + 28.1 = 186.1%. The average is 186.1 / 10 = 18.61%.\",\n \"To find the median, list the percentages in order and identify the middle number. The ordered percentages are 5.9, 9.7, 11.8, 13.9, 14.2, 15.6, 19.7, 28.1, 29.4, 37.8. The middle values are 14.2 and 15.6. The median is the average of these two: (14.2 + 15.6) / 2 = 14.9%.\"\n ],\n \"answer\": [\n \"Protected Land Percentage by Country\",\n \"Germany\",\n \"28.1%\",\n \"3\",\n \"China\",\n \"57.5%\",\n \"India\",\n \"Canada, 2.1%\",\n \"18.61%\",\n \"14.9%\"\n ]\n}"},"metadata":{"kind":"string","value":"{\n \"figure_type\": \"Funnel Chart\",\n \"persona\": \"A geography professor at ISU who inspires students with a passion for environmental conservation and mapping technologies\",\n \"topic\": \"Percentage of protected land areas across different countries\"\n}"},"data":{"kind":"string","value":"Country,Protected Land Percentage\nBrazil,29.4\nUnited States,13.9\nAustralia,19.7\nChina,15.6\nCanada,11.8\nRussia,9.7\nIndia,5.9\nMexico,14.2\nGermany,37.8\nFrance,28.1"},"code":{"kind":"string","value":"import plotly.graph_objects as go\nimport pandas as pd\nimport io\nfrom PIL import Image\n\ndef generate_plot(df):\n # Sort the dataframe by Protected Land Percentage in descending order\n df_sorted = df.sort_values('Protected Land Percentage', ascending=False)\n\n # Create the funnel chart\n fig = go.Figure(go.Funnel(\n y=df_sorted['Country'],\n x=df_sorted['Protected Land Percentage'],\n textposition=\"inside\",\n textinfo=\"value+percent initial\",\n opacity=0.65,\n marker={\"color\": [\"#66c2a5\", \"#fc8d62\", \"#8da0cb\", \"#e78ac3\", \"#a6d854\",\n \"#ffd92f\", \"#e5c494\", \"#b3b3b3\", \"#7fc97f\", \"#beaed4\"],\n \"line\": {\"width\": [4, 2, 2, 3, 1, 1, 1, 1, 1, 1],\n \"color\": [\"#66c2a5\", \"#fc8d62\", \"#8da0cb\", \"#e78ac3\", \"#a6d854\",\n \"#ffd92f\", \"#e5c494\", \"#b3b3b3\", \"#7fc97f\", \"#beaed4\"]}},\n connector={\"line\": {\"color\": \"royalblue\", \"dash\": \"dot\", \"width\": 3}}\n ))\n\n # Update the layout\n fig.update_layout(\n title={\n 'text': \"Protected Land Percentage by Country\",\n 'y':0.95,\n 'x':0.5,\n 'xanchor': 'center',\n 'yanchor': 'top',\n 'font': dict(size=24, color='royalblue', family=\"Arial Black\")\n },\n font=dict(family=\"Arial\", size=14, color=\"royalblue\"),\n paper_bgcolor='lightskyblue',\n plot_bgcolor='aliceblue',\n showlegend=False,\n width=900,\n height=700,\n margin=dict(l=120, r=20, t=100, b=50)\n )\n\n # Add a annotation\n fig.add_annotation(\n text=\"Data source: Environmental Protection Agency\",\n xref=\"paper\", yref=\"paper\",\n x=0.5, y=-0.1,\n showarrow=False,\n font=dict(size=12, color=\"gray\"),\n align=\"center\",\n )\n\n # Save the plot to a BytesIO object\n img_bytes = io.BytesIO()\n fig.write_image(img_bytes, format=\"png\", scale=2)\n img_bytes.seek(0)\n \n # Return the plot as a PIL Image object\n return Image.open(img_bytes)"}}},{"rowIdx":21,"cells":{"id":{"kind":"string","value":"MatplotlibChartPipeline_stem_2-8"},"image":{"kind":"image","value":{"src":"https://datasets-server.huggingface.co/assets/allenai/CoSyn-400K/--/86e46e1fd5e754d056169f0fb38f06c6997ff7de/--/chart/train/21/image/image.jpg?Expires=1741922929&Signature=iv7CXqQSYNvwzw6GpC8A~xy-JO5CeQS1ZZgMSULAcoBbvr6x~AodLO2rWvRxP-kMoK21JGtQpwLwOefLq1byCAuS01OhqDQnQc4pFCZ6TnKydfJjTPigaxhJ8hzXkPUTYyp4vt1XFsz41mRUJksVvon0x~UObkbbyu8ZB9wFQaf3byRbxbKJvmdjAodDVTrC1JEMMOz6jHi9cyvLN-PZT-pAWZVjCibVq6buEK0a2fuvXPXWCKHTJbCczxWG0pBCQHFausarh8nxwvtZNGWDsewEIUaVul8dkWpqsDCi1UPXztN62X3xyizY1SZEjezEXuLAuHGYPRYlWHej5a3CzA__&Key-Pair-Id=K3EI6M078Z3AC3","height":1440,"width":1396},"inferredFromUrl":false},"qa_pairs":{"kind":"list like","value":{"question":["What is the title of the subplot collection?","How does the rainfall in the Amazon Rainforest in January 2023 compare to January 1970?","Which region has the lowest rainfall in July?","What is the overall trend in rainfall for California from January to December?","How many regions are depicted in the subplots?","In which months does the Sahara Desert receive the highest rainfall?","What is the total annual rainfall in the Amazon Rainforest in 2023?","What is the difference in rainfall between California and the Amazon Rainforest in June?","Which region has the greatest variability in monthly rainfall?","Is there any region where rainfall remains constant throughout the year?"],"explanation":["The title is displayed at the top and summarizes the content of the figure by mentioning the topic of rainfall distribution and climate change impact.","The subplot for the Amazon Rainforest indicates the monthly rainfall for both 2023 and 1970. For January, we can see that the rainfall was 450 mm in 2023 and 520 mm in 1970.","By comparing the July data across the subplots, the Sahara Desert shows the lowest rainfall with 0 mm marked for July.","Observing the California subplot, the rainfall starts moderately high in January, decreases to almost nil around July and August, then gradually increases back towards December.","The figure has subplots for each unique region mentioned. Counting the titles within the subplots, we find four regions: Amazon Rainforest, Sahara Desert, California, and Amazon Rainforest (1970).","Reviewing the subplot for the Sahara Desert, the months of February, April, and November each have approximately 3 mm of rainfall, the highest for this region.","Summing up the monthly values in the subplot for the Amazon Rainforest (2023): 450 + 470 + 460 + 480 + 500 + 520 + 510 + 530 + 550 + 560 + 540 + 530 = 6100 mm.","According to their respective subplots, California has 10 mm of rainfall in June while the Amazon Rainforest has 520 mm. The difference is 520 - 10 = 510 mm.","By observing the fluctuation within each subplot, the Amazon Rainforest and its 1970 counterpart show a much larger monthly range compared to California and the Sahara Desert, whose variability is quite small.","By checking each subplot, no region shows a perfectly constant rainfall across all months; there are variations in each of the regions depicted."],"answer":["Monthly Rainfall Distribution in Various Regions","The rainfall is lower in 2023","Sahara Desert","Decreasing then increasing","Four regions","February, April, November","6100 mm","510 mm","Amazon Rainforest","No"]},"string":"{\n \"question\": [\n \"What is the title of the subplot collection?\",\n \"How does the rainfall in the Amazon Rainforest in January 2023 compare to January 1970?\",\n \"Which region has the lowest rainfall in July?\",\n \"What is the overall trend in rainfall for California from January to December?\",\n \"How many regions are depicted in the subplots?\",\n \"In which months does the Sahara Desert receive the highest rainfall?\",\n \"What is the total annual rainfall in the Amazon Rainforest in 2023?\",\n \"What is the difference in rainfall between California and the Amazon Rainforest in June?\",\n \"Which region has the greatest variability in monthly rainfall?\",\n \"Is there any region where rainfall remains constant throughout the year?\"\n ],\n \"explanation\": [\n \"The title is displayed at the top and summarizes the content of the figure by mentioning the topic of rainfall distribution and climate change impact.\",\n \"The subplot for the Amazon Rainforest indicates the monthly rainfall for both 2023 and 1970. For January, we can see that the rainfall was 450 mm in 2023 and 520 mm in 1970.\",\n \"By comparing the July data across the subplots, the Sahara Desert shows the lowest rainfall with 0 mm marked for July.\",\n \"Observing the California subplot, the rainfall starts moderately high in January, decreases to almost nil around July and August, then gradually increases back towards December.\",\n \"The figure has subplots for each unique region mentioned. Counting the titles within the subplots, we find four regions: Amazon Rainforest, Sahara Desert, California, and Amazon Rainforest (1970).\",\n \"Reviewing the subplot for the Sahara Desert, the months of February, April, and November each have approximately 3 mm of rainfall, the highest for this region.\",\n \"Summing up the monthly values in the subplot for the Amazon Rainforest (2023): 450 + 470 + 460 + 480 + 500 + 520 + 510 + 530 + 550 + 560 + 540 + 530 = 6100 mm.\",\n \"According to their respective subplots, California has 10 mm of rainfall in June while the Amazon Rainforest has 520 mm. The difference is 520 - 10 = 510 mm.\",\n \"By observing the fluctuation within each subplot, the Amazon Rainforest and its 1970 counterpart show a much larger monthly range compared to California and the Sahara Desert, whose variability is quite small.\",\n \"By checking each subplot, no region shows a perfectly constant rainfall across all months; there are variations in each of the regions depicted.\"\n ],\n \"answer\": [\n \"Monthly Rainfall Distribution in Various Regions\",\n \"The rainfall is lower in 2023\",\n \"Sahara Desert\",\n \"Decreasing then increasing\",\n \"Four regions\",\n \"February, April, November\",\n \"6100 mm\",\n \"510 mm\",\n \"Amazon Rainforest\",\n \"No\"\n ]\n}"},"metadata":{"kind":"string","value":"{\n \"figure_type\": \"subplots of stem plot\",\n \"persona\": \"A creative storyteller who relies on the colorist's work to evoke the right emotions in the readers\",\n \"topic\": \"Monthly Rainfall Distribution in Various Regions to Depict Climate Change Impact\"\n}"},"data":{"kind":"string","value":"Region,Month,Rainfall (mm)\nAmazon Rainforest,January,450\nAmazon Rainforest,February,470\nAmazon Rainforest,March,460\nAmazon Rainforest,April,480\nAmazon Rainforest,May,500\nAmazon Rainforest,June,520\nAmazon Rainforest,July,510\nAmazon Rainforest,August,530\nAmazon Rainforest,September,550\nAmazon Rainforest,October,560\nAmazon Rainforest,November,540\nAmazon Rainforest,December,530\nSahara Desert,January,2\nSahara Desert,February,3\nSahara Desert,March,1\nSahara Desert,April,3\nSahara Desert,May,2\nSahara Desert,June,1\nSahara Desert,July,0\nSahara Desert,August,1\nSahara Desert,September,2\nSahara Desert,October,2\nSahara Desert,November,3\nSahara Desert,December,1\nCalifornia,January,80\nCalifornia,February,70\nCalifornia,March,90\nCalifornia,April,60\nCalifornia,May,30\nCalifornia,June,10\nCalifornia,July,5\nCalifornia,August,5\nCalifornia,September,10\nCalifornia,October,20\nCalifornia,November,60\nCalifornia,December,80\nAmazon Rainforest (1970),January,520\nAmazon Rainforest (1970),February,540\nAmazon Rainforest (1970),March,530\nAmazon Rainforest (1970),April,550\nAmazon Rainforest (1970),May,570\nAmazon Rainforest (1970),June,590\nAmazon Rainforest (1970),July,580\nAmazon Rainforest (1970),August,600\nAmazon Rainforest (1970),September,620\nAmazon Rainforest (1970),October,630\nAmazon Rainforest (1970),November,610\nAmazon Rainforest (1970),December,600"},"code":{"kind":"string","value":"import numpy as np\nimport matplotlib.pyplot as plt\nfrom io import BytesIO\nfrom PIL import Image\n\ndef generate_plot(data):\n plt.style.use('tableau-colorblind10')\n \n regions = data['Region'].unique()\n n_regions = len(regions)\n \n fig, axs = plt.subplots(2, 2, figsize=(16, 16))\n axs = axs.flatten()\n \n months = ['Jan', 'Feb', 'Mar', 'Apr', 'May', 'Jun', 'Jul', 'Aug', 'Sep', 'Oct', 'Nov', 'Dec']\n \n for i, region in enumerate(regions):\n region_data = data[data['Region'] == region]\n rainfall = region_data['Rainfall (mm)'].values\n \n markerline, stemlines, baseline = axs[i].stem(months, rainfall, linefmt='--', markerfmt='o', bottom=0)\n \n plt.setp(markerline, markersize=10, markeredgewidth=2, markeredgecolor='black', markerfacecolor='white')\n plt.setp(stemlines, linewidth=2, color='darkgrey', alpha=0.7)\n plt.setp(baseline, linewidth=2, color='black')\n \n axs[i].set_title(region, fontsize=18, fontweight='bold', pad=20)\n axs[i].set_ylabel('Rainfall (mm)', fontsize=14)\n axs[i].set_ylim(bottom=0)\n \n if region == 'Sahara Desert':\n axs[i].set_ylim(top=5)\n elif region == 'California':\n axs[i].set_ylim(top=100)\n else:\n axs[i].set_ylim(top=700)\n \n axs[i].tick_params(axis='both', which='major', labelsize=12)\n axs[i].grid(True, linestyle=':', alpha=0.6)\n \n for j, rainfall in enumerate(region_data['Rainfall (mm)']):\n axs[i].annotate(f'{rainfall}', (j, rainfall), xytext=(0, 5), \n textcoords='offset points', ha='center', va='bottom',\n fontsize=10, fontweight='bold')\n \n fig.suptitle('Monthly Rainfall Distribution in Various Regions', fontsize=24, fontweight='bold', y=1.02)\n fig.text(0.5, 0.01, 'Month', ha='center', fontsize=16)\n \n plt.tight_layout()\n \n buffer = BytesIO()\n plt.savefig(buffer, format='png', dpi=300, bbox_inches='tight')\n buffer.seek(0)\n image = Image.open(buffer)\n \n return image"}}},{"rowIdx":22,"cells":{"id":{"kind":"string","value":"HTMLChartPipeline_bar_2-447"},"image":{"kind":"image","value":{"src":"https://datasets-server.huggingface.co/assets/allenai/CoSyn-400K/--/86e46e1fd5e754d056169f0fb38f06c6997ff7de/--/chart/train/22/image/image.jpg?Expires=1741922929&Signature=u01fWdA62EeNblFm8gqm-RzTPGsUA5XVRM8jozwjP6gNkABPmUsJwHRZYUjrNFz-brrd6iOH~1zuiAuO9k-cyW31oEaoPecCOi41XrgIZzo9R2H2fbNREuVMtfWMErxybDGO2kj2ZZVZBe2gtf-bGR-qCRq1OKb38LaRU0-h-iz8totCjrzRKQJKMluunxwzNfB7JKWijdCnaPTNpcCLVLn1NlWdBzy2MvYRhQLYwGXxXlnwPJ-PzpiJniDnzc75H337p6cLMR6UJr1Q7fiONLkPnhN76YX9lOsp1Y4IvPIJBHFTbeU-3FuOal5IHI96NirrvcBQaR8rF09It6nHDw__&Key-Pair-Id=K3EI6M078Z3AC3","height":800,"width":861},"inferredFromUrl":false},"qa_pairs":{"kind":"list like","value":{"question":["Which subject is the most favored among students with high proficiency in math?","What is the difference in the percentage of students favoring science between high and low proficiency groups?","Which subject has the closest percentage of favorability to math among students with low proficiency in math?","Rank the favorability of English across proficiency levels from highest to lowest.","Compare the favorability of art and music among medium proficiency students. Which is favored more and by how much?","What is the average percentage of students favoring computer science across all proficiency levels?","Which subject has the least favorability among students with high proficiency in math, and what is the percentage?","By how much does the favorability of physical education differ between high and medium proficiency groups?","How does the favorability of history compare between high and low proficiency groups?"],"explanation":["The chart shows the percentage of students with different proficiency levels in math who favor each subject. For students with high proficiency, the highest bar indicates the most favored subject, which is \"Math\" at 75%.","From the chart, the percentage of students favoring science is 70% for high proficiency and 30% for low proficiency. The difference is calculated as 70 - 30 = 40.","For the low proficiency group, the favorability percentage of math is 20%. By comparing other subjects, the closest is \"Music\" at 25%, which is only 5% different.","For English, the percentages are: Medium (60%), Low (55%), and High (50%). Thus, the ranking from highest to lowest is Medium, Low, High.","For students with medium proficiency, the favorability of art is 30% and for music, it is 15%. Art is favored more, and the difference is 30 - 15 = 15.","The percentages for computer science are High (60%), Medium (45%), and Low (30%). Summing them up, 60 + 45 + 30 = 135. Dividing by the number of groups: 135 / 3 = 45.","The subject with the smallest percentage bar for students with high proficiency is \"Music,\" with a favorability of 10%.","The chart shows that for physical education, high proficiency students have a favorability of 25% and medium proficiency students have 35%. The difference is 35 - 25 = 10.","For history, high proficiency students have a favorability of 30% while low proficiency students have 50%. Low proficiency favor history more by 20%."],"answer":["Math","40","Music","Medium, Low, High","Art; 15","45","Music; 10","10","Low proficiency; 20"]},"string":"{\n \"question\": [\n \"Which subject is the most favored among students with high proficiency in math?\",\n \"What is the difference in the percentage of students favoring science between high and low proficiency groups?\",\n \"Which subject has the closest percentage of favorability to math among students with low proficiency in math?\",\n \"Rank the favorability of English across proficiency levels from highest to lowest.\",\n \"Compare the favorability of art and music among medium proficiency students. Which is favored more and by how much?\",\n \"What is the average percentage of students favoring computer science across all proficiency levels?\",\n \"Which subject has the least favorability among students with high proficiency in math, and what is the percentage?\",\n \"By how much does the favorability of physical education differ between high and medium proficiency groups?\",\n \"How does the favorability of history compare between high and low proficiency groups?\"\n ],\n \"explanation\": [\n \"The chart shows the percentage of students with different proficiency levels in math who favor each subject. For students with high proficiency, the highest bar indicates the most favored subject, which is \\\"Math\\\" at 75%.\",\n \"From the chart, the percentage of students favoring science is 70% for high proficiency and 30% for low proficiency. The difference is calculated as 70 - 30 = 40.\",\n \"For the low proficiency group, the favorability percentage of math is 20%. By comparing other subjects, the closest is \\\"Music\\\" at 25%, which is only 5% different.\",\n \"For English, the percentages are: Medium (60%), Low (55%), and High (50%). Thus, the ranking from highest to lowest is Medium, Low, High.\",\n \"For students with medium proficiency, the favorability of art is 30% and for music, it is 15%. Art is favored more, and the difference is 30 - 15 = 15.\",\n \"The percentages for computer science are High (60%), Medium (45%), and Low (30%). Summing them up, 60 + 45 + 30 = 135. Dividing by the number of groups: 135 / 3 = 45.\",\n \"The subject with the smallest percentage bar for students with high proficiency is \\\"Music,\\\" with a favorability of 10%.\",\n \"The chart shows that for physical education, high proficiency students have a favorability of 25% and medium proficiency students have 35%. The difference is 35 - 25 = 10.\",\n \"For history, high proficiency students have a favorability of 30% while low proficiency students have 50%. Low proficiency favor history more by 20%.\"\n ],\n \"answer\": [\n \"Math\",\n \"40\",\n \"Music\",\n \"Medium, Low, High\",\n \"Art; 15\",\n \"45\",\n \"Music; 10\",\n \"10\",\n \"Low proficiency; 20\"\n ]\n}"},"metadata":{"kind":"string","value":"{\n \"figure_type\": \"horizontal bar chart\",\n \"persona\": \"A high school student struggling to understand trigonometry and looking for extra help in math\",\n \"topic\": \"Favorite School Subjects Among High School Students with Different Proficiency Levels in Math\"\n}"},"data":{"kind":"string","value":"{\n \"FavoriteSubjects\": [\n {\n \"ProficiencyLevel\": \"High\",\n \"Math\": 75,\n \"Science\": 70,\n \"English\": 50,\n \"History\": 30,\n \"PhysicalEducation\": 25,\n \"Art\": 20,\n \"Music\": 10,\n \"ComputerScience\": 60\n },\n {\n \"ProficiencyLevel\": \"Medium\",\n \"Math\": 50,\n \"Science\": 55,\n \"English\": 60,\n \"History\": 40,\n \"PhysicalEducation\": 35,\n \"Art\": 30,\n \"Music\": 15,\n \"ComputerScience\": 45\n },\n {\n \"ProficiencyLevel\": \"Low\",\n \"Math\": 20,\n \"Science\": 30,\n \"English\": 55,\n \"History\": 50,\n \"PhysicalEducation\": 45,\n \"Art\": 40,\n \"Music\": 25,\n \"ComputerScience\": 30\n }\n ]\n}"},"code":{"kind":"string","value":"\n\n\n \n \n \n \n \n\n\n \n\n \n\n"}}},{"rowIdx":23,"cells":{"id":{"kind":"string","value":"PlotlyChartPipeline_sub_1-231"},"image":{"kind":"image","value":{"src":"https://datasets-server.huggingface.co/assets/allenai/CoSyn-400K/--/86e46e1fd5e754d056169f0fb38f06c6997ff7de/--/chart/train/23/image/image.jpg?Expires=1741922929&Signature=T7AbTRQA2K70n9hip9uwjDdD98MjMEZkyaZ~FzAjQQaqhpTNcmivb4aVcYpJpM9Qf6~uaZ1SxK66~leAG4zCLPPWyhdggBJltxMtgStb1BwsFuCt2Bwgo9AsYb9HN5PLNsnmp9KGgOTq2CSPY-JBJGNygJTh1RYzfVLch-B9q5~6Uan3~BpYLRaZcmZ7DPuYbry8D9Gge1cwHUr87SQXbdrlfIVsKZXR9L~ATJFPGPQuzwg0W2KWoCwHdzigNTt9gAfQQ7k5itMOMSdTjcPrGi-jvxT1zpX3cEzdT-asQALxCC7vh8bd45A2xGv66ndWO0RBaL72NeiLkmOEibWNxg__&Key-Pair-Id=K3EI6M078Z3AC3","height":1800,"width":1600},"inferredFromUrl":false},"qa_pairs":{"kind":"list like","value":{"question":["what is the title of the figure?","how many regions are displayed in the figure?","which region has the highest hourly rate for senior-level legal coaches?","what is the difference in hourly rates between entry-level and senior-level legal coaches in the Midwest?","which region has the lowest hourly rate for mid-career legal coaches?","what is the average hourly rate for entry-level legal coaches across all regions?","how many subplots are there in the figure?","which region shows the largest increase in hourly rates from entry-level to mid-career?","what is the median hourly rate for senior-level legal coaches across all regions?"],"explanation":["The title of a figure is usually placed at the top and gives an overall description of the displayed data. In this case, it reads \"Comparison of Hourly Rates for Legal Coaches Across US Regions\".","Each subplot (Entry-Level, Mid-Career, Senior-Level) has bars representing different regions in the US. Counting the bars or regions' names, we see there are 6 regions represented.","Look at the third subplot labeled \"Senior-Level\". Identify the tallest bar, which is for the \"West Coast\" with an hourly rate of 325.","Locate the bars for the Midwest in both the Entry-Level and Senior-Level subplots. The values are 100 for Entry-Level and 250 for Senior-Level. The difference is 250 - 100 = 150.","In the Mid-Career subplot, identify the shortest bar, which is for the \"Midwest\" with an hourly rate of 175.","Add the hourly rates for Entry-Level across all regions: 150 (Northeast) + 125 (Southeast) + 100 (Midwest) + 130 (Southwest) + 175 (West Coast) + 120 (Rocky Mountain) = 800. Divide by the number of regions, 800/6 = 133.33.","By visual inspection, there are three labeled subplots: Entry-Level, Mid-Career, and Senior-Level.","Calculate the increase for each region from Entry-Level to Mid-Career:\n- Northeast: 225 - 150 = 75\n- Southeast: 200 - 125 = 75\n- Midwest: 175 - 100 = 75\n- Southwest: 210 - 130 = 80\n- West Coast: 250 - 175 = 75\n- Rocky Mountain: 190 - 120 = 70\nThe largest increase is in the Southwest with 80.","List the hourly rates for Senior-Level legal coaches across all regions: 300, 275, 250, 280, 325, 260. Order them: 250, 260, 275, 280, 300, 325. The median value (middle value) is the average of the third and fourth values: (275+280)/2 = 277.5."],"answer":["Comparison of Hourly Rates for Legal Coaches Across US Regions","6","West Coast","150","Midwest","133.33","3","Southwest","277.5"]},"string":"{\n \"question\": [\n \"what is the title of the figure?\",\n \"how many regions are displayed in the figure?\",\n \"which region has the highest hourly rate for senior-level legal coaches?\",\n \"what is the difference in hourly rates between entry-level and senior-level legal coaches in the Midwest?\",\n \"which region has the lowest hourly rate for mid-career legal coaches?\",\n \"what is the average hourly rate for entry-level legal coaches across all regions?\",\n \"how many subplots are there in the figure?\",\n \"which region shows the largest increase in hourly rates from entry-level to mid-career?\",\n \"what is the median hourly rate for senior-level legal coaches across all regions?\"\n ],\n \"explanation\": [\n \"The title of a figure is usually placed at the top and gives an overall description of the displayed data. In this case, it reads \\\"Comparison of Hourly Rates for Legal Coaches Across US Regions\\\".\",\n \"Each subplot (Entry-Level, Mid-Career, Senior-Level) has bars representing different regions in the US. Counting the bars or regions' names, we see there are 6 regions represented.\",\n \"Look at the third subplot labeled \\\"Senior-Level\\\". Identify the tallest bar, which is for the \\\"West Coast\\\" with an hourly rate of 325.\",\n \"Locate the bars for the Midwest in both the Entry-Level and Senior-Level subplots. The values are 100 for Entry-Level and 250 for Senior-Level. The difference is 250 - 100 = 150.\",\n \"In the Mid-Career subplot, identify the shortest bar, which is for the \\\"Midwest\\\" with an hourly rate of 175.\",\n \"Add the hourly rates for Entry-Level across all regions: 150 (Northeast) + 125 (Southeast) + 100 (Midwest) + 130 (Southwest) + 175 (West Coast) + 120 (Rocky Mountain) = 800. Divide by the number of regions, 800/6 = 133.33.\",\n \"By visual inspection, there are three labeled subplots: Entry-Level, Mid-Career, and Senior-Level.\",\n \"Calculate the increase for each region from Entry-Level to Mid-Career:\\n- Northeast: 225 - 150 = 75\\n- Southeast: 200 - 125 = 75\\n- Midwest: 175 - 100 = 75\\n- Southwest: 210 - 130 = 80\\n- West Coast: 250 - 175 = 75\\n- Rocky Mountain: 190 - 120 = 70\\nThe largest increase is in the Southwest with 80.\",\n \"List the hourly rates for Senior-Level legal coaches across all regions: 300, 275, 250, 280, 325, 260. Order them: 250, 260, 275, 280, 300, 325. The median value (middle value) is the average of the third and fourth values: (275+280)/2 = 277.5.\"\n ],\n \"answer\": [\n \"Comparison of Hourly Rates for Legal Coaches Across US Regions\",\n \"6\",\n \"West Coast\",\n \"150\",\n \"Midwest\",\n \"133.33\",\n \"3\",\n \"Southwest\",\n \"277.5\"\n ]\n}"},"metadata":{"kind":"string","value":"{\n \"figure_type\": \"vertical subplots\",\n \"persona\": \"A former trial attorney who has transitioned into a career as a legal skills coach and mentor\",\n \"topic\": \"Comparison of hourly rates for legal coaches across different regions in the US\"\n}"},"data":{"kind":"string","value":"Region,Entry-Level,Mid-Career,Senior-Level\nNortheast,150,225,300\nSoutheast,125,200,275\nMidwest,100,175,250\nSouthwest,130,210,280\nWest Coast,175,250,325\nRocky Mountain,120,190,260"},"code":{"kind":"string","value":"import plotly.graph_objects as go\nfrom plotly.subplots import make_subplots\nimport pandas as pd\nimport io\nfrom PIL import Image\n\ndef generate_plot(data):\n # Create subplots\n fig = make_subplots(rows=3, cols=1, subplot_titles=(\"Entry-Level\", \"Mid-Career\", \"Senior-Level\"))\n\n # Custom color palette\n colors = ['#1f77b4', '#ff7f0e', '#2ca02c', '#d62728', '#9467bd', '#8c564b']\n\n # Add traces for each level\n for i, level in enumerate(['Entry-Level', 'Mid-Career', 'Senior-Level']):\n fig.add_trace(\n go.Bar(\n x=data['Region'],\n y=data[level],\n marker_color=colors,\n text=data[level],\n textposition='outside',\n textfont=dict(size=10),\n hovertemplate='%{x}: $%{y}/hour'\n ),\n row=i+1, col=1\n )\n\n # Update layout\n fig.update_layout(\n title_text=\"Comparison of Hourly Rates for Legal Coaches Across US Regions\",\n title_font=dict(size=24),\n height=900,\n width=800,\n showlegend=False,\n plot_bgcolor='rgba(240,240,240,0.95)',\n paper_bgcolor='rgba(240,240,240,0.95)',\n margin=dict(l=50, r=50, t=100, b=50),\n )\n\n # Update axes\n for i in range(1, 4):\n fig.update_xaxes(\n tickangle=45,\n title_text=\"Region\",\n title_font=dict(size=12),\n row=i, col=1\n )\n fig.update_yaxes(\n title_text=\"Hourly Rate ($)\",\n title_font=dict(size=12),\n range=[0, max(data.iloc[:, 1:].max()) * 1.1], # Set y-axis range\n row=i, col=1\n )\n\n # Save plot to BytesIO object\n img_bytes = io.BytesIO()\n fig.write_image(img_bytes, format=\"png\", scale=2)\n img_bytes.seek(0)\n \n # Convert to PIL Image\n img = Image.open(img_bytes)\n \n return img"}}},{"rowIdx":24,"cells":{"id":{"kind":"string","value":"MatplotlibChartPipeline_stair_1-282"},"image":{"kind":"image","value":{"src":"https://datasets-server.huggingface.co/assets/allenai/CoSyn-400K/--/86e46e1fd5e754d056169f0fb38f06c6997ff7de/--/chart/train/24/image/image.jpg?Expires=1741922929&Signature=b3vknslo2itGmHeWo3JksibL8bJ8-aSNhZhx0P3tMgDKMplcKURXFfRq8T6jASkDqkv-tFcbervOMtOZ2AAxODRU1Pmf4H-MdTGn-GK~2vwZdWCSlZWMrDRfz-V6UT1KcpCyQsAhc-nU6LQsRJAUzzf8tO7aDgd4H6XThtlTlLya9c8dLJN1uDx2c3lYLm3Xtk~ilwzXlnJ1MDMAdULLR-taC5c~qTJhCXunFKhDI6fxhsgiPVdmYwGMTfj1pOLPA8qaKvbjr30oSv-iDrvBDJULVsArxG1b9D004BOfmVq8Gl2iCLcFKKA0~LPOV6CFvIImYVzod5qKG54Dg90STg__&Key-Pair-Id=K3EI6M078Z3AC3","height":1440,"width":2169},"inferredFromUrl":false},"qa_pairs":{"kind":"list like","value":{"question":["What's the title of the figure?","What is the percentage of voters supporting Pro-Security Policies in 1980?","By how much did the percentage of voters for Pacifist Policies increase from 1985 to 1990?","Which year had the largest gap between Pro-Security Policies and Pacifist Policies?","How many years show that at least 60% of voters supported Pro-Security Policies?","Which year marks a notable dip in support for Pro-Security Policies compared to previous years?","What is the trend for Pacifist Policies from 1995 to 2005?","How did voter attitudes toward Pro-Security Policies change from 2015 to 2023?","What is the average percentage of Pro-Security Policies across the given years?","In which years did the proportions of Pro-Security and Pacifist Policies voters add up to exactly 100%?"],"explanation":["The title is located at the top of the chart and usually provides a summary of what the figure represents. By looking at the figure, the title reads \"Historical Shifts in Voter Attitudes Toward National Security Policies\".","To find this, look at the year 1980 on the x-axis and then follow the stair step up to the Pro-Security Policies line. The value at that point is labeled directly on the plot.","First, identify the percentage for Pacifist Policies in 1985 and 1990 by looking at the plot. In 1985, it is 25%, and in 1990, it is 45%. Subtract the 1985 percentage from the 1990 percentage (45% - 25%).","To determine this, examine the differences between the two series for each year and identify the year with the maximum difference. The largest gap is in 1985, with Pro-Security at 75% and Pacifist at 25%, thus a gap of 50%.","Count the years where the percentage value for Pro-Security Policies is equal to or exceeds 60%. These years are 1970, 1980, 1985, 2000, 2010, and 2015.","Identify a significant drop in support levels by comparing each year's percentage to the previous one. The notable dip occurs in 1990, where the support for Pro-Security Policies drops sharply from 75% in 1985 to 55%.","Check the stair plot section from 1995 to 2005. The percentages go from 48% in 1995 to 42% in 2005, showing a decreasing trend.","Examine the percentages for Pro-Security Policies in 2015, 2020, and 2023. They are 67%, 64%, and 61%, respectively, indicating a decreasing trend.","Sum all the Pro-Security percentages (65 + 62 + 70 + 75 + 55 + 52 + 60 + 58 + 63 + 67 + 64 + 61) and then divide by the number of years (12). The calculation is (752/12).","All the given years sum up to 100% because the only categories displayed are Pro-Security and Pacifist Policies, representing binary choices so their complements add up to 100%. Check the entire data set to confirm this."],"answer":["Historical Shifts in Voter Attitudes Toward National Security Policies","70%","20%","1985","6 years","1990","Decreasing","Decreased","62.67%","All years"]},"string":"{\n \"question\": [\n \"What's the title of the figure?\",\n \"What is the percentage of voters supporting Pro-Security Policies in 1980?\",\n \"By how much did the percentage of voters for Pacifist Policies increase from 1985 to 1990?\",\n \"Which year had the largest gap between Pro-Security Policies and Pacifist Policies?\",\n \"How many years show that at least 60% of voters supported Pro-Security Policies?\",\n \"Which year marks a notable dip in support for Pro-Security Policies compared to previous years?\",\n \"What is the trend for Pacifist Policies from 1995 to 2005?\",\n \"How did voter attitudes toward Pro-Security Policies change from 2015 to 2023?\",\n \"What is the average percentage of Pro-Security Policies across the given years?\",\n \"In which years did the proportions of Pro-Security and Pacifist Policies voters add up to exactly 100%?\"\n ],\n \"explanation\": [\n \"The title is located at the top of the chart and usually provides a summary of what the figure represents. By looking at the figure, the title reads \\\"Historical Shifts in Voter Attitudes Toward National Security Policies\\\".\",\n \"To find this, look at the year 1980 on the x-axis and then follow the stair step up to the Pro-Security Policies line. The value at that point is labeled directly on the plot.\",\n \"First, identify the percentage for Pacifist Policies in 1985 and 1990 by looking at the plot. In 1985, it is 25%, and in 1990, it is 45%. Subtract the 1985 percentage from the 1990 percentage (45% - 25%).\",\n \"To determine this, examine the differences between the two series for each year and identify the year with the maximum difference. The largest gap is in 1985, with Pro-Security at 75% and Pacifist at 25%, thus a gap of 50%.\",\n \"Count the years where the percentage value for Pro-Security Policies is equal to or exceeds 60%. These years are 1970, 1980, 1985, 2000, 2010, and 2015.\",\n \"Identify a significant drop in support levels by comparing each year's percentage to the previous one. The notable dip occurs in 1990, where the support for Pro-Security Policies drops sharply from 75% in 1985 to 55%.\",\n \"Check the stair plot section from 1995 to 2005. The percentages go from 48% in 1995 to 42% in 2005, showing a decreasing trend.\",\n \"Examine the percentages for Pro-Security Policies in 2015, 2020, and 2023. They are 67%, 64%, and 61%, respectively, indicating a decreasing trend.\",\n \"Sum all the Pro-Security percentages (65 + 62 + 70 + 75 + 55 + 52 + 60 + 58 + 63 + 67 + 64 + 61) and then divide by the number of years (12). The calculation is (752/12).\",\n \"All the given years sum up to 100% because the only categories displayed are Pro-Security and Pacifist Policies, representing binary choices so their complements add up to 100%. Check the entire data set to confirm this.\"\n ],\n \"answer\": [\n \"Historical Shifts in Voter Attitudes Toward National Security Policies\",\n \"70%\",\n \"20%\",\n \"1985\",\n \"6 years\",\n \"1990\",\n \"Decreasing\",\n \"Decreased\",\n \"62.67%\",\n \"All years\"\n ]\n}"},"metadata":{"kind":"string","value":"{\n \"figure_type\": \"stair plot\",\n \"persona\": \"A hardline political pundit who often debates against pacifist policies on their radio show\",\n \"topic\": \"Historical Shifts in Voter Attitudes Toward National Security Policies from 1970 to Present\"\n}"},"data":{"kind":"string","value":"Year,Pro-Security Policies,Pacifist Policies\n1970,65,35\n1975,62,38\n1980,70,30\n1985,75,25\n1990,55,45\n1995,52,48\n2000,60,40\n2005,58,42\n2010,63,37\n2015,67,33\n2020,64,36\n2023,61,39"},"code":{"kind":"string","value":"import numpy as np\nimport matplotlib.pyplot as plt\nfrom io import BytesIO\nfrom PIL import Image\n\ndef generate_plot(df):\n plt.style.use('default')\n \n fig, ax = plt.subplots(figsize=(12, 8))\n \n years = df['Year'].values\n pro_security = df['Pro-Security Policies'].values\n pacifist = df['Pacifist Policies'].values\n \n ax.step(years, pro_security, where='post', label='Pro-Security Policies', \n color='#D62728', linewidth=3, marker='o', markersize=8)\n ax.step(years, pacifist, where='post', label='Pacifist Policies', \n color='#2CA02C', linewidth=3, marker='s', markersize=8)\n \n ax.fill_between(years, pro_security, step=\"post\", alpha=0.3, color='#D62728')\n ax.fill_between(years, pacifist, step=\"post\", alpha=0.3, color='#2CA02C')\n \n ax.set_xlim(1970, 2023)\n ax.set_ylim(0, 100)\n ax.set_xticks(years)\n ax.set_xticklabels(years, rotation=45, ha='right')\n \n ax.set_title(\"Historical Shifts in Voter Attitudes\\nToward National Security Policies\", \n fontsize=20, fontweight='bold', pad=20)\n ax.set_xlabel(\"Year\", fontsize=14, fontweight='bold', labelpad=10)\n ax.set_ylabel(\"Percentage of Voters (%)\", fontsize=14, fontweight='bold', labelpad=10)\n \n ax.grid(True, linestyle='--', alpha=0.7)\n ax.legend(loc='center left', bbox_to_anchor=(1, 0.5), fontsize=12)\n \n ax.spines['top'].set_visible(False)\n ax.spines['right'].set_visible(False)\n \n for i, (pro, pac) in enumerate(zip(pro_security, pacifist)):\n ax.annotate(f'{pro}%', (years[i], pro), xytext=(0, 5), \n textcoords='offset points', ha='center', va='bottom', \n fontsize=10, fontweight='bold', color='#D62728')\n ax.annotate(f'{pac}%', (years[i], pac), xytext=(0, -15), \n textcoords='offset points', ha='center', va='top', \n fontsize=10, fontweight='bold', color='#2CA02C')\n \n plt.tight_layout()\n \n buf = BytesIO()\n plt.savefig(buf, format='png', dpi=300, bbox_inches='tight')\n buf.seek(0)\n img = Image.open(buf)\n \n return img"}}},{"rowIdx":25,"cells":{"id":{"kind":"string","value":"VegaLiteChartPipeline_bubble_1-325"},"image":{"kind":"image","value":{"src":"https://datasets-server.huggingface.co/assets/allenai/CoSyn-400K/--/86e46e1fd5e754d056169f0fb38f06c6997ff7de/--/chart/train/25/image/image.jpg?Expires=1741922929&Signature=dm0hlG4EcjLwIKHf~pEXlT6EbrmaYihfbsqcjqcezpDAAcvG6IBvVPUllxlUTZkNyJBVG~qIb1laPCaQESHYJLSEcwPB~Ch~asaP9vWqXVQBFqozTDCVUFzXVWtEGwvfIIqrgjjvimesM~KzqrDd4S~X8HijRpJeIy4NEhZO01-6zMaW63b~HhEu0wBGTP6S2asywnxjZhjcSZFNtAeA6qPKvIRN1aeNb4OPZzgyxSdoesJNxrDdvxPCzdJq3diO7Rl1Bv1-f1DPMCuuU3VUqakzg~bcK5koFVvXQvWDJFvTnmLHQmALtv8ahYyE2iKWlCMRTOqFT9Qe~mL-~xnK-w__&Key-Pair-Id=K3EI6M078Z3AC3","height":1428,"width":2277},"inferredFromUrl":false},"qa_pairs":{"kind":"list like","value":{"question":["What is the title of the chart?","Which city has the highest percentage of public transportation usage?","What is the carbon emissions metric tons per capita for Los Angeles?","Which city has the lowest carbon emissions per capita?","What is the relationship between public transportation usage and carbon emissions for the cities shown?","Which city has the largest bubble size, and what does it represent?","Compare the public transportation usage between San Francisco and Miami.","Which city has higher carbon emissions per capita, Seattle or Denver?","What is the average carbon emissions per capita for New York City and San Francisco combined?","How does Boston's public transportation usage compare to Chicago's?"],"explanation":["The title of the chart is displayed at the top and describes the chart's content.","By looking at the x-axis, we find the bubble that is furthest to the right. This bubble corresponds to New York City.","Identify the bubble for Los Angeles by its color and tooltip, then read its position on the y-axis.","By locating the bubble closest to the bottom of the y-axis and checking the tooltip or color, we find it corresponds to Boston.","The chart generally shows that higher public transportation usage tends to relate to lower carbon emissions, as seen in cities like New York City and San Francisco, whereas cities like Houston, with low public transportation usage, have higher emissions.","The largest bubble is identified by comparing bubble sizes. New York City has the largest bubble size, which represents an additional dimension of interest, such as population or another relevant metric.","Locate the bubbles for San Francisco and Miami, then compare their positions on the x-axis to determine their public transportation usage percentages.","By comparing the y-axis positions of Seattle and Denver's bubbles, we determine which one is higher.","Sum the emissions for both cities (5.5 + 4.2) and divide by 2 to get the average.","Locate the bubbles for Boston and Chicago, then compare their positions on the x-axis to determine their public transportation usage percentages."],"answer":["Public Transportation Usage vs. Carbon Emissions","New York City","8.2","Boston","Inverse relationship","New York City","San Francisco has 38%, and Miami has 18%","Denver","4.85","Boston has 34% and Chicago has 32%"]},"string":"{\n \"question\": [\n \"What is the title of the chart?\",\n \"Which city has the highest percentage of public transportation usage?\",\n \"What is the carbon emissions metric tons per capita for Los Angeles?\",\n \"Which city has the lowest carbon emissions per capita?\",\n \"What is the relationship between public transportation usage and carbon emissions for the cities shown?\",\n \"Which city has the largest bubble size, and what does it represent?\",\n \"Compare the public transportation usage between San Francisco and Miami.\",\n \"Which city has higher carbon emissions per capita, Seattle or Denver?\",\n \"What is the average carbon emissions per capita for New York City and San Francisco combined?\",\n \"How does Boston's public transportation usage compare to Chicago's?\"\n ],\n \"explanation\": [\n \"The title of the chart is displayed at the top and describes the chart's content.\",\n \"By looking at the x-axis, we find the bubble that is furthest to the right. This bubble corresponds to New York City.\",\n \"Identify the bubble for Los Angeles by its color and tooltip, then read its position on the y-axis.\",\n \"By locating the bubble closest to the bottom of the y-axis and checking the tooltip or color, we find it corresponds to Boston.\",\n \"The chart generally shows that higher public transportation usage tends to relate to lower carbon emissions, as seen in cities like New York City and San Francisco, whereas cities like Houston, with low public transportation usage, have higher emissions.\",\n \"The largest bubble is identified by comparing bubble sizes. New York City has the largest bubble size, which represents an additional dimension of interest, such as population or another relevant metric.\",\n \"Locate the bubbles for San Francisco and Miami, then compare their positions on the x-axis to determine their public transportation usage percentages.\",\n \"By comparing the y-axis positions of Seattle and Denver's bubbles, we determine which one is higher.\",\n \"Sum the emissions for both cities (5.5 + 4.2) and divide by 2 to get the average.\",\n \"Locate the bubbles for Boston and Chicago, then compare their positions on the x-axis to determine their public transportation usage percentages.\"\n ],\n \"answer\": [\n \"Public Transportation Usage vs. Carbon Emissions\",\n \"New York City\",\n \"8.2\",\n \"Boston\",\n \"Inverse relationship\",\n \"New York City\",\n \"San Francisco has 38%, and Miami has 18%\",\n \"Denver\",\n \"4.85\",\n \"Boston has 34% and Chicago has 32%\"\n ]\n}"},"metadata":{"kind":"string","value":"{\n \"figure_type\": \"subplot of Bubble Chart\",\n \"persona\": \"A former colleague of Senator Ken Donnelly who deeply respected his contributions to environmental initiatives\",\n \"topic\": \"Public Transportation Usage vs. Carbon Emissions across Metropolitan Areas\"\n}"},"data":{"kind":"string","value":"City,Public_Transportation_Usage_Percent,Carbon_Emissions_Metric_Tons_Per_Capita,Bubble_Size\nNew York City,56.0,5.5,8.4\nLos Angeles,26.0,8.2,3.9\nChicago,32.0,6.0,2.7\nSan Francisco,38.0,4.2,0.9\nBoston,34.0,4.0,0.7\nWashington D.C.,37.0,5.9,0.7\nSeattle,23.0,4.5,0.7\nDenver,15.0,6.7,0.7\nMiami,18.0,6.5,0.5\nHouston,5.0,10.2,2.3"},"code":{"kind":"string","value":"{\n \"$schema\": \"https://vega.github.io/schema/vega-lite/v5.json\",\n \"width\": 600,\n \"height\": 400,\n \"title\": {\n \"text\": \"Public Transportation Usage vs. Carbon Emissions\",\n \"fontSize\": 20,\n \"font\": \"Helvetica\",\n \"color\": \"#3D3D3D\"\n },\n \"data\": {\n \"values\": [\n {\"City\": \"New York City\", \"Public_Transportation_Usage_Percent\": 56.0, \"Carbon_Emissions_Metric_Tons_Per_Capita\": 5.5, \"Bubble_Size\": 8.4},\n {\"City\": \"Los Angeles\", \"Public_Transportation_Usage_Percent\": 26.0, \"Carbon_Emissions_Metric_Tons_Per_Capita\": 8.2, \"Bubble_Size\": 3.9},\n {\"City\": \"Chicago\", \"Public_Transportation_Usage_Percent\": 32.0, \"Carbon_Emissions_Metric_Tons_Per_Capita\": 6.0, \"Bubble_Size\": 2.7},\n {\"City\": \"San Francisco\", \"Public_Transportation_Usage_Percent\": 38.0, \"Carbon_Emissions_Metric_Tons_Per_Capita\": 4.2, \"Bubble_Size\": 0.9},\n {\"City\": \"Boston\", \"Public_Transportation_Usage_Percent\": 34.0, \"Carbon_Emissions_Metric_Tons_Per_Capita\": 4.0, \"Bubble_Size\": 0.7},\n {\"City\": \"Washington D.C.\", \"Public_Transportation_Usage_Percent\": 37.0, \"Carbon_Emissions_Metric_Tons_Per_Capita\": 5.9, \"Bubble_Size\": 0.7},\n {\"City\": \"Seattle\", \"Public_Transportation_Usage_Percent\": 23.0, \"Carbon_Emissions_Metric_Tons_Per_Capita\": 4.5, \"Bubble_Size\": 0.7},\n {\"City\": \"Denver\", \"Public_Transportation_Usage_Percent\": 15.0, \"Carbon_Emissions_Metric_Tons_Per_Capita\": 6.7, \"Bubble_Size\": 0.7},\n {\"City\": \"Miami\", \"Public_Transportation_Usage_Percent\": 18.0, \"Carbon_Emissions_Metric_Tons_Per_Capita\": 6.5, \"Bubble_Size\": 0.5},\n {\"City\": \"Houston\", \"Public_Transportation_Usage_Percent\": 5.0, \"Carbon_Emissions_Metric_Tons_Per_Capita\": 10.2, \"Bubble_Size\": 2.3}\n ]\n },\n \"mark\": {\n \"type\": \"circle\",\n \"opacity\": 0.7\n },\n \"encoding\": {\n \"x\": {\n \"field\": \"Public_Transportation_Usage_Percent\",\n \"type\": \"quantitative\",\n \"title\": \"Public Transportation Usage (%)\",\n \"axis\": {\n \"titleFont\": \"Helvetica\",\n \"titleFontSize\": 14,\n \"labelFont\": \"Helvetica\",\n \"labelFontSize\": 12\n }\n },\n \"y\": {\n \"field\": \"Carbon_Emissions_Metric_Tons_Per_Capita\",\n \"type\": \"quantitative\",\n \"title\": \"Carbon Emissions (Metric Tons Per Capita)\",\n \"axis\": {\n \"titleFont\": \"Helvetica\",\n \"titleFontSize\": 14,\n \"labelFont\": \"Helvetica\",\n \"labelFontSize\": 12\n }\n },\n \"size\": {\n \"field\": \"Bubble_Size\",\n \"type\": \"quantitative\",\n \"scale\": {\n \"range\": [100, 2000]\n },\n \"legend\": null\n },\n \"color\": {\n \"field\": \"City\",\n \"type\": \"nominal\",\n \"scale\": {\n \"scheme\": \"category10\"\n },\n \"legend\": {\n \"orient\": \"right\",\n \"title\": \"City\",\n \"titleFont\": \"Helvetica\",\n \"titleFontSize\": 12,\n \"labelFont\": \"Helvetica\",\n \"labelFontSize\": 10\n }\n },\n \"tooltip\": [\n {\"field\": \"City\", \"type\": \"nominal\"},\n {\"field\": \"Public_Transportation_Usage_Percent\", \"type\": \"quantitative\", \"format\": \".1f\", \"title\": \"Public Transportation Usage (%)\"},\n {\"field\": \"Carbon_Emissions_Metric_Tons_Per_Capita\", \"type\": \"quantitative\", \"format\": \".1f\", \"title\": \"Carbon Emissions (Metric Tons Per Capita)\"}\n ]\n },\n \"config\": {\n \"view\": {\"stroke\": null},\n \"background\": \"#F5F5F5\"\n }\n}"}}},{"rowIdx":26,"cells":{"id":{"kind":"string","value":"MatplotlibChart_basic_6-376"},"image":{"kind":"image","value":{"src":"https://datasets-server.huggingface.co/assets/allenai/CoSyn-400K/--/86e46e1fd5e754d056169f0fb38f06c6997ff7de/--/chart/train/26/image/image.jpg?Expires=1741922929&Signature=at~bWXwp05qa~TyiQBoohKn7eSgxMgrjgWPfSZNbmsPcfjIueAz~fOlumpmO~dhXLganNoXZxfcSn11mal8FyMOrH-lc8vY3qVilgjyMKBxsawii-mkHwqSqKhyZQb332eGgxm7o6npoNHs9Zp-rv326NNaQlcGkqp0PedkC~J1iigI9J~5L9XJCmwUcy9shTKh37XJOsC4t4dJtQAjvaCgbaaXRZsPdPVQ8geRGYEVfp9nPq61h5-dqFbUtIpT3BRwbAzuup~xGogvqIKgzIyWVlxYUbe8O~Ze2wK49bNnCyiXNt-nEEQaDp4lBcmR0OL-nNch~RIPbtVzDa9-BOA__&Key-Pair-Id=K3EI6M078Z3AC3","height":1440,"width":1804},"inferredFromUrl":false},"qa_pairs":{"kind":"list like","value":{"question":["What crop has the highest water usage?","Which crop has the largest yield per hectare?","Between Potato and Tomato, which has a higher water usage efficiency?","Which crop has the smallest bubble size?","Arrange Wheat, Maize, and Rice in descending order of their waste usage.","What crop with a water usage below 5000 m3/ha has the highest yield?","Which crop has both water usage and yield above 4000 m3/ha and 30 ton/ha respectively?","What is the average yield of crops that use less than 3500 m3/ha of water?","Identify the crop with the lowest yield but still having a bubble size equal or more than 50.","If we increase the water usage in Rice to 8000 m3/ha, how does it compare to the bubble size of Cotton?"],"explanation":["To determine the crop with the highest water usage, we need to look at the x-axis values. The crop with the highest x-value corresponds to the highest water usage. From the data, it is clear that Sugarcane has the highest water usage, at 10,000 m3/ha.","We need to identify the crop with the highest y-axis value, as the y-axis represents the yield per hectare. From the chart, Sugarcane has the largest yield per hectare at 80 ton/ha.","The water usage efficiency is calculated as yield divided by water usage. From the data, Potato has an efficiency of 0.00667 ton/m3 and Tomato has an efficiency of 0.00825 ton/m3. Comparing these values, Tomato has a higher water usage efficiency.","Bubble size correlates with the numerical values given in the BubbleSize column. The smallest bubble size in the data is 15, which corresponds to Spinach and Lettuce.","We must look at the x-axis values for Wheat, Maize, and Rice. Wheat has 3500 m3/ha, Maize has 5000 m3/ha, and Rice has 7000 m3/ha. Arranging in descending order: Rice, Maize, Wheat.","We focus on crops with water usage less than 5000 m3/ha, then look for the highest y-axis value among them. From the chart, crops such as Onion (25 ton/ha), Spinach (15 ton/ha), and Lettuce (20 ton/ha) need to be checked. The crop with the highest yield is Onion at 25 ton/ha.","From the datasets, we need to check for conditions where both water usage and yield are above the mentioned values. Carrot, Cabbage, and Sugarcane meet these criteria.","The crops are Peas (3 ton/ha), Lettuce (20 ton/ha), Spinach (15 ton/ha), Onion (25 ton/ha). The average yield is calculated as (3 + 20 + 15 + 25) / 4 = 15.75 ton/ha.","Within the dataset, we look for crops with a bubble size of 50 or above: Maize, Cotton, Potato, Sugarcane. Cotton yield is 3.2 ton/ha, the lowest yield among them.","Initially, Cotton has a water usage of 7500 m3/ha, which is slightly below the new Rice water usage. The comparison is between 8000 m3/ha for Rice and 7500 m3/ha for Cotton."],"answer":["Sugarcane","Sugarcane","Tomato","Spinach, Lettuce","Rice, Maize, Wheat","Onion","Carrot, Cabbage, Sugarcane","15.75","Cotton","Rice uses more"]},"string":"{\n \"question\": [\n \"What crop has the highest water usage?\",\n \"Which crop has the largest yield per hectare?\",\n \"Between Potato and Tomato, which has a higher water usage efficiency?\",\n \"Which crop has the smallest bubble size?\",\n \"Arrange Wheat, Maize, and Rice in descending order of their waste usage.\",\n \"What crop with a water usage below 5000 m3/ha has the highest yield?\",\n \"Which crop has both water usage and yield above 4000 m3/ha and 30 ton/ha respectively?\",\n \"What is the average yield of crops that use less than 3500 m3/ha of water?\",\n \"Identify the crop with the lowest yield but still having a bubble size equal or more than 50.\",\n \"If we increase the water usage in Rice to 8000 m3/ha, how does it compare to the bubble size of Cotton?\"\n ],\n \"explanation\": [\n \"To determine the crop with the highest water usage, we need to look at the x-axis values. The crop with the highest x-value corresponds to the highest water usage. From the data, it is clear that Sugarcane has the highest water usage, at 10,000 m3/ha.\",\n \"We need to identify the crop with the highest y-axis value, as the y-axis represents the yield per hectare. From the chart, Sugarcane has the largest yield per hectare at 80 ton/ha.\",\n \"The water usage efficiency is calculated as yield divided by water usage. From the data, Potato has an efficiency of 0.00667 ton/m3 and Tomato has an efficiency of 0.00825 ton/m3. Comparing these values, Tomato has a higher water usage efficiency.\",\n \"Bubble size correlates with the numerical values given in the BubbleSize column. The smallest bubble size in the data is 15, which corresponds to Spinach and Lettuce.\",\n \"We must look at the x-axis values for Wheat, Maize, and Rice. Wheat has 3500 m3/ha, Maize has 5000 m3/ha, and Rice has 7000 m3/ha. Arranging in descending order: Rice, Maize, Wheat.\",\n \"We focus on crops with water usage less than 5000 m3/ha, then look for the highest y-axis value among them. From the chart, crops such as Onion (25 ton/ha), Spinach (15 ton/ha), and Lettuce (20 ton/ha) need to be checked. The crop with the highest yield is Onion at 25 ton/ha.\",\n \"From the datasets, we need to check for conditions where both water usage and yield are above the mentioned values. Carrot, Cabbage, and Sugarcane meet these criteria.\",\n \"The crops are Peas (3 ton/ha), Lettuce (20 ton/ha), Spinach (15 ton/ha), Onion (25 ton/ha). The average yield is calculated as (3 + 20 + 15 + 25) / 4 = 15.75 ton/ha.\",\n \"Within the dataset, we look for crops with a bubble size of 50 or above: Maize, Cotton, Potato, Sugarcane. Cotton yield is 3.2 ton/ha, the lowest yield among them.\",\n \"Initially, Cotton has a water usage of 7500 m3/ha, which is slightly below the new Rice water usage. The comparison is between 8000 m3/ha for Rice and 7500 m3/ha for Cotton.\"\n ],\n \"answer\": [\n \"Sugarcane\",\n \"Sugarcane\",\n \"Tomato\",\n \"Spinach, Lettuce\",\n \"Rice, Maize, Wheat\",\n \"Onion\",\n \"Carrot, Cabbage, Sugarcane\",\n \"15.75\",\n \"Cotton\",\n \"Rice uses more\"\n ]\n}"},"metadata":{"kind":"string","value":"{\n \"figure_type\": \"bubble chart\",\n \"persona\": \"A local farmer from a different cultural background who appreciates the Surinamese farmer's expertise and seeks their guidance in sustainable farming practices\",\n \"topic\": \"Water Usage Efficiency Across Different Crops\"\n}"},"data":{"kind":"string","value":"Crop,Water Usage (m3/ha),Yield (ton/ha),Efficiency (ton/m3),BubbleSize\nWheat,3500,6.5,0.00186,30\nMaize,5000,8.2,0.00164,50\nRice,7000,5.0,0.00071,70\nSoybean,3200,2.5,0.00078,25\nCotton,7500,3.2,0.00043,50\nTomato,4000,33.0,0.00825,33\nPotato,6000,40.0,0.00667,60\nOnion,3000,25.0,0.00833,20\nLettuce,2500,20.0,0.008,15\nSunflower,4000,2.0,0.0005,30\nSugarcane,10000,80.0,0.008,80\nBarley,3600,5.5,0.00153,35\nPeas,2500,3.0,0.0012,20\nCarrot,3000,30.0,0.01,30\nSpinach,1500,15.0,0.01,15\nCabbage,4500,45.0,0.01,40"},"code":{"kind":"string","value":"import numpy as np\nimport matplotlib.pyplot as plt\nimport pandas as pd\nfrom io import BytesIO\nfrom PIL import Image\n\ndef generate_plot(df):\n plt.style.use('seaborn-v0_8-muted')\n \n # Define figure and axis\n fig, ax = plt.subplots(figsize=(10, 8))\n \n # Scatter plot\n sc = ax.scatter(df['Water Usage (m3/ha)'], df['Yield (ton/ha)'], \n s=df['BubbleSize'] * 10, # Scale bubble sizes\n c=np.random.rand(len(df), 3), # Random colors for each bubble\n alpha=0.6, edgecolors=\"w\", linewidth=0.5, marker='o')\n \n # Title and labels\n ax.set_title('Water Usage Efficiency Across Different Crops', fontsize=16, fontweight='bold')\n ax.set_xlabel('Water Usage (m3/ha)', fontsize=14)\n ax.set_ylabel('Yield (ton/ha)', fontsize=14)\n \n # Annotate each point\n for i, row in df.iterrows():\n ax.annotate(row['Crop'], \n (row['Water Usage (m3/ha)'], row['Yield (ton/ha)']),\n fontsize=10, ha='center')\n \n # Set axis limits\n ax.set_xlim(0, df['Water Usage (m3/ha)'].max()*1.1)\n ax.set_ylim(0, df['Yield (ton/ha)'].max()*1.1)\n \n # Create grid\n ax.grid(True, linestyle='--', alpha=0.5)\n \n # Save to BytesIO\n buf = BytesIO()\n plt.tight_layout()\n plt.savefig(buf, format='png', bbox_inches='tight', dpi=300)\n buf.seek(0)\n img = Image.open(buf)\n \n return img"}}},{"rowIdx":27,"cells":{"id":{"kind":"string","value":"MatplotlibChart_pie_claude_1-1136"},"image":{"kind":"image","value":{"src":"https://datasets-server.huggingface.co/assets/allenai/CoSyn-400K/--/86e46e1fd5e754d056169f0fb38f06c6997ff7de/--/chart/train/27/image/image.jpg?Expires=1741922929&Signature=cF4jKlSAIZf0rxe5EhQa5OjfabKngNVluBZ2bwzETuqqVYBYMXm1LhhIkuhxRAaRoc3kGI~2HOzdNHFdUHaKma9i1z7ehZ6VS67X2agZCVH5N65clttB6xuhhwg4-Hqlo9b1ucmLkk5nwIyz1oTxWJEKyE1hysa93RQpEUNkQ0mgoSK4fmmjIBOYQTn6Y~zjPn29IABtsraKDDDWFzB-KUA7EYdRWJLuFFUZXM0Mi2nxSnCtkDJGhpYL88zk2YCEkCBrfBKGskr6sqwJ9oY9n2H725uS~sFibFDU~lT11Yv3J3FKvZXb0H7OYMFNpaRvtnwkVtNOx5rMzdfT4SYDiA__&Key-Pair-Id=K3EI6M078Z3AC3","height":1440,"width":2204},"inferredFromUrl":false},"qa_pairs":{"kind":"list like","value":{"question":["What percentage of the construction projects are residential?","Which project type has the smallest percentage?","How much larger is the percentage of commercial projects compared to industrial projects?","What is the combined percentage of public infrastructure and industrial projects?","Which project type is more common, residential or public infrastructure?","Are there more commercial projects than public infrastructure projects?","What is the percentage difference between the largest and smallest project types?","What is the average percentage of the mixed-use development, public infrastructure, and industrial projects?","Which color represents the public infrastructure projects?","Is the percentage of residential projects greater than the combined percentage of industrial and mixed-use development projects?"],"explanation":["To find the percentage of residential construction projects, refer to the pie chart section labeled \"Residential.\" The percentage is clearly marked on the chart.","Compare all the sections of the pie chart and identify the one with the smallest percentage figure. The smallest section is indicated.","Locate the percentages for both the commercial and industrial sections on the pie chart. Subtract the industrial percentage from the commercial percentage: 28% - 10% = 18%.","Add the percentages of public infrastructure and industrial sections: 22% + 10% = 32%.","Compare the residential and public infrastructure sections of the pie chart. The residential section has a higher percentage than the public infrastructure section (35% vs. 22%).","Compare the commercial and public infrastructure sections of the pie chart. The commercial section has 28%, which is more than the 22% of the public infrastructure section.","Identify the largest and smallest project percentages: residential (35%) and mixed-use development (5%). Subtract the smallest from the largest: 35% - 5% = 30%.","Add the percentages of mixed-use development, public infrastructure, and industrial projects: 5% + 22% + 10% = 37%. Divide by the number of project types (3): 37% / 3 ≈ 12.33%.","Refer to the visual color key in the pie chart and match the color to the label \"Public Infrastructure.\"","Compare the residential percentage (35%) with the combined percentage of industrial (10%) and mixed-use development (5%): 10% + 5% = 15%. Since 35% > 15%, the answer is yes."],"answer":["35%","Mixed-Use Development","18%","32%","Residential","Yes","30%","12.33%","Orange","Yes"]},"string":"{\n \"question\": [\n \"What percentage of the construction projects are residential?\",\n \"Which project type has the smallest percentage?\",\n \"How much larger is the percentage of commercial projects compared to industrial projects?\",\n \"What is the combined percentage of public infrastructure and industrial projects?\",\n \"Which project type is more common, residential or public infrastructure?\",\n \"Are there more commercial projects than public infrastructure projects?\",\n \"What is the percentage difference between the largest and smallest project types?\",\n \"What is the average percentage of the mixed-use development, public infrastructure, and industrial projects?\",\n \"Which color represents the public infrastructure projects?\",\n \"Is the percentage of residential projects greater than the combined percentage of industrial and mixed-use development projects?\"\n ],\n \"explanation\": [\n \"To find the percentage of residential construction projects, refer to the pie chart section labeled \\\"Residential.\\\" The percentage is clearly marked on the chart.\",\n \"Compare all the sections of the pie chart and identify the one with the smallest percentage figure. The smallest section is indicated.\",\n \"Locate the percentages for both the commercial and industrial sections on the pie chart. Subtract the industrial percentage from the commercial percentage: 28% - 10% = 18%.\",\n \"Add the percentages of public infrastructure and industrial sections: 22% + 10% = 32%.\",\n \"Compare the residential and public infrastructure sections of the pie chart. The residential section has a higher percentage than the public infrastructure section (35% vs. 22%).\",\n \"Compare the commercial and public infrastructure sections of the pie chart. The commercial section has 28%, which is more than the 22% of the public infrastructure section.\",\n \"Identify the largest and smallest project percentages: residential (35%) and mixed-use development (5%). Subtract the smallest from the largest: 35% - 5% = 30%.\",\n \"Add the percentages of mixed-use development, public infrastructure, and industrial projects: 5% + 22% + 10% = 37%. Divide by the number of project types (3): 37% / 3 ≈ 12.33%.\",\n \"Refer to the visual color key in the pie chart and match the color to the label \\\"Public Infrastructure.\\\"\",\n \"Compare the residential percentage (35%) with the combined percentage of industrial (10%) and mixed-use development (5%): 10% + 5% = 15%. Since 35% > 15%, the answer is yes.\"\n ],\n \"answer\": [\n \"35%\",\n \"Mixed-Use Development\",\n \"18%\",\n \"32%\",\n \"Residential\",\n \"Yes\",\n \"30%\",\n \"12.33%\",\n \"Orange\",\n \"Yes\"\n ]\n}"},"metadata":{"kind":"string","value":"{\n \"figure_type\": \"pie chart\",\n \"persona\": \"A construction contractor who works on various infrastructure projects in the same city\",\n \"topic\": \"Distribution of construction project types in the city (residential, commercial, industrial, public infrastructure)\"\n}"},"data":{"kind":"string","value":"Project Type,Percentage\nResidential,35\nCommercial,28\nPublic Infrastructure,22\nIndustrial,10\nMixed-Use Development,5"},"code":{"kind":"string","value":"import numpy as np\nimport matplotlib.pyplot as plt\nfrom io import BytesIO\nfrom PIL import Image\n\ndef generate_plot(data):\n plt.style.use('seaborn-v0_8-poster')\n \n # Prepare data\n project_types = data['Project Type']\n percentages = data['Percentage']\n \n # Create custom color palette\n colors = ['#FF9999', '#66B2FF', '#99FF99', '#FFCC99', '#FF99CC']\n \n # Create the pie chart\n fig, ax = plt.subplots(figsize=(12, 8))\n wedges, texts, autotexts = ax.pie(percentages, labels=project_types, autopct='%1.1f%%', \n startangle=90, colors=colors, \n wedgeprops=dict(width=0.6, edgecolor='white'))\n \n # Customize the appearance\n plt.setp(autotexts, size=12, weight=\"bold\", color=\"white\")\n plt.setp(texts, size=14, weight=\"bold\")\n \n # Add a title\n ax.set_title(\"Distribution of Construction Project Types\", fontsize=24, pad=20)\n \n # Add a legend\n ax.legend(wedges, project_types,\n title=\"Project Types\",\n loc=\"center left\",\n bbox_to_anchor=(1, 0, 0.5, 1))\n \n # Save the plot to BytesIO\n buf = BytesIO()\n plt.savefig(buf, format='png', dpi=300, bbox_inches='tight')\n buf.seek(0)\n \n # Convert to PIL Image\n img = Image.open(buf)\n \n return img"}}},{"rowIdx":28,"cells":{"id":{"kind":"string","value":"PlotlyChartPipeline_icicle_2-42"},"image":{"kind":"image","value":{"src":"https://datasets-server.huggingface.co/assets/allenai/CoSyn-400K/--/86e46e1fd5e754d056169f0fb38f06c6997ff7de/--/chart/train/28/image/image.jpg?Expires=1741922929&Signature=FJrh~ulYWKTZsOEBZQCmefp~ayKk4qF9loXHI6n71KrVqdehTVqBq3xrUh0ZOLSntlZAEYWf~EWyr3d4lIgSPMk4kStuCgYibBW15Pt~kqqfBkvhrxGnT-ZP5HiIo9nyiEf1NDoVhRKP6K7XePYnop7dXgPckOpYvBlRVReUJcizHwwhXvDTZn5NiyQFsL6FHJzwkJWuh96n856Q-JYyX41kJP1iVuXO~S~F8Vs~0d46TTK93Ww~OfzRrM3XNOM3cix90icNo2kmdYvJOyr8Yu03VXYps7uaU5jNlQpFSZhv8s1LR83vfolwSk~w5W21jSM5EyZxFZDltTBxMeuwVQ__&Key-Pair-Id=K3EI6M078Z3AC3","height":1600,"width":2000},"inferredFromUrl":false},"qa_pairs":{"kind":"list like","value":{"question":["How many main stages are there in the chart?","Which society is associated with Slash-and-Burn Horticultural?","What are the two sub-stages under the Agrarian stage?","Compare the activities associated with the Bedouin and Trobriand Islanders societies. How do they differ?","Which stage has societies practicing both subsistence farming and mass production?","List all societies under the Post-Industrial stage and their respective examples.","Which sub-stage under the Industrial stage represents early industrialization, and which society is an example?","How does the Trobriand Islanders' example of human activity differ from that of the Yanomami?","What example of human activity is associated with the United States in the chart?"],"explanation":["The chart lists different social and cultural evolution stages, and by counting the main categories, we find five: Hunting-Gathering, Horticultural, Pastoral, Agrarian, Industrial, and Post-Industrial.","By looking at the Icicle Chart, the Yanomami society is classified under the Slash-and-Burn sub-stage within the Horticultural stage.","The chart displays the sub-stages of the Agrarian stage by expanding it: Subsistence and Feudal.","The Bedouin society, under the Pastoral stage, Semi-Nomadic sub-stage, practices goat and camel herding with seasonal settlements, while the Trobriand Islanders, under Intensive Horticultural, focus on yam cultivation and complex exchange systems. These activities differ significantly: the former centers around herding, while the latter involves intensive agriculture and trade.","The Agrarian stage includes subsistence farming by the Aztecs, while the Industrial stage includes mass production by the United States. Additionally, by comparing these stages, we see both practices are represented across these distinct stages.","Under the Post-Industrial stage, there are two societies: Japan with robotics and high-tech manufacturing, and Singapore with financial services and knowledge-based industries.","The early sub-stage of the Industrial stage is represented by Victorian England, which is associated with the factory system and urbanization.","The Trobriand Islanders practice yam cultivation and complex exchange systems under Intensive Horticultural, while the Yanomami practice shifting cultivation in the Amazon rainforest under Slash-and-Burn Horticultural. Shifting cultivation involves moving from one land patch to another, while intensive cultivation and trade involve more permanent agricultural processes and social structures.","The Icicle Chart shows the United States under the Late Industrial stage with mass production and consumerism as the associated activities."],"answer":["6","Yanomami","Subsistence and Feudal","Bedouin: Goat and camel herding; Trobriand Islanders: Yam cultivation and complex exchange","Agrarian and Industrial","Japan: robotics and high-tech manufacturing; Singapore: financial services and knowledge-based industries","Early: Victorian England","Intensive cultivation and trade vs. shifting cultivation","Mass production and consumerism"]},"string":"{\n \"question\": [\n \"How many main stages are there in the chart?\",\n \"Which society is associated with Slash-and-Burn Horticultural?\",\n \"What are the two sub-stages under the Agrarian stage?\",\n \"Compare the activities associated with the Bedouin and Trobriand Islanders societies. How do they differ?\",\n \"Which stage has societies practicing both subsistence farming and mass production?\",\n \"List all societies under the Post-Industrial stage and their respective examples.\",\n \"Which sub-stage under the Industrial stage represents early industrialization, and which society is an example?\",\n \"How does the Trobriand Islanders' example of human activity differ from that of the Yanomami?\",\n \"What example of human activity is associated with the United States in the chart?\"\n ],\n \"explanation\": [\n \"The chart lists different social and cultural evolution stages, and by counting the main categories, we find five: Hunting-Gathering, Horticultural, Pastoral, Agrarian, Industrial, and Post-Industrial.\",\n \"By looking at the Icicle Chart, the Yanomami society is classified under the Slash-and-Burn sub-stage within the Horticultural stage.\",\n \"The chart displays the sub-stages of the Agrarian stage by expanding it: Subsistence and Feudal.\",\n \"The Bedouin society, under the Pastoral stage, Semi-Nomadic sub-stage, practices goat and camel herding with seasonal settlements, while the Trobriand Islanders, under Intensive Horticultural, focus on yam cultivation and complex exchange systems. These activities differ significantly: the former centers around herding, while the latter involves intensive agriculture and trade.\",\n \"The Agrarian stage includes subsistence farming by the Aztecs, while the Industrial stage includes mass production by the United States. Additionally, by comparing these stages, we see both practices are represented across these distinct stages.\",\n \"Under the Post-Industrial stage, there are two societies: Japan with robotics and high-tech manufacturing, and Singapore with financial services and knowledge-based industries.\",\n \"The early sub-stage of the Industrial stage is represented by Victorian England, which is associated with the factory system and urbanization.\",\n \"The Trobriand Islanders practice yam cultivation and complex exchange systems under Intensive Horticultural, while the Yanomami practice shifting cultivation in the Amazon rainforest under Slash-and-Burn Horticultural. Shifting cultivation involves moving from one land patch to another, while intensive cultivation and trade involve more permanent agricultural processes and social structures.\",\n \"The Icicle Chart shows the United States under the Late Industrial stage with mass production and consumerism as the associated activities.\"\n ],\n \"answer\": [\n \"6\",\n \"Yanomami\",\n \"Subsistence and Feudal\",\n \"Bedouin: Goat and camel herding; Trobriand Islanders: Yam cultivation and complex exchange\",\n \"Agrarian and Industrial\",\n \"Japan: robotics and high-tech manufacturing; Singapore: financial services and knowledge-based industries\",\n \"Early: Victorian England\",\n \"Intensive cultivation and trade vs. shifting cultivation\",\n \"Mass production and consumerism\"\n ]\n}"},"metadata":{"kind":"string","value":"{\n \"figure_type\": \"Icicle Chart\",\n \"persona\": \"A member of an academic forum dedicated to discussing anthropology theories and fieldwork experiences\",\n \"topic\": \"Socio-cultural evolution stages with examples from different societies\"\n}"},"data":{"kind":"string","value":"Stage,Sub-Stage,Society,Example\nHunting-Gathering,Nomadic,!Kung San,Seasonal migration for food sources\nHunting-Gathering,Semi-Sedentary,Inuit,Winter settlements and summer hunting camps\nHorticultural,Slash-and-Burn,Yanomami,Shifting cultivation in Amazon rainforest\nHorticultural,Intensive,Trobriand Islanders,Yam cultivation and complex exchange systems\nPastoral,Nomadic,Maasai,Cattle herding across East African savannas\nPastoral,Semi-Nomadic,Bedouin,Goat and camel herding with seasonal settlements\nAgrarian,Subsistence,Aztec,Chinampas farming and tribute system\nAgrarian,Feudal,Medieval Europe,Manor system and hierarchical social structure\nIndustrial,Early,Victorian England,Factory system and urbanization\nIndustrial,Late,United States,Mass production and consumerism\nPost-Industrial,Information Age,Japan,Robotics and high-tech manufacturing\nPost-Industrial,Service Economy,Singapore,Financial services and knowledge-based industries"},"code":{"kind":"string","value":"import pandas as pd\nimport plotly.express as px\nfrom io import BytesIO\nfrom PIL import Image\n\ndef generate_plot(data):\n # Prepare the data\n data['value'] = 1 # Add a value column for sizing\n\n # Create the icicle chart\n fig = px.icicle(data, path=['Stage', 'Sub-Stage', 'Society', 'Example'], values='value',\n color='Stage', \n color_discrete_sequence=px.colors.qualitative.Pastel,\n title='Socio-cultural Evolution Stages',\n hover_data=['Example'])\n\n # Customize the layout\n fig.update_layout(\n font_family=\"Courier New\",\n font_size=12,\n title_font_size=24,\n title_font_family=\"Georgia\",\n title_x=0.5,\n width=1000,\n height=800,\n margin=dict(t=50, l=25, r=25, b=25)\n )\n\n # Customize the traces\n fig.update_traces(\n textinfo=\"label+text\",\n text=data['Example'],\n hovertemplate='%{label}
Example: %{customdata[0]}'\n )\n\n # Save the plot to BytesIO\n img_bytes = BytesIO()\n fig.write_image(img_bytes, format=\"png\", scale=2)\n img_bytes.seek(0)\n \n # Return as PIL Image\n return Image.open(img_bytes)"}}},{"rowIdx":29,"cells":{"id":{"kind":"string","value":"MatplotlibChart_basic_4-712"},"image":{"kind":"image","value":{"src":"https://datasets-server.huggingface.co/assets/allenai/CoSyn-400K/--/86e46e1fd5e754d056169f0fb38f06c6997ff7de/--/chart/train/29/image/image.jpg?Expires=1741922929&Signature=os~~934TmRL8ETLxwpddkvl42lepeO103wRFzixOc8~pBom3Pdj2KNZDjL6a5WhdoAVIX0-a7u6PCuuxbyei-qmq41lyOO8mqEJKDz0EkALWzARvYw5wE2qXRRoSsXHSPwIJl9eMWoMi9an1B0~99VmhtGyzn0Qyg8ffJyBORObQjmHJPM2FmqHBA7zxzfMkUrvDyKd3knwTIymR5oo78XRb-LmXg-N2zcQ8ICn2f31kztGzjhxm7DcFcSmPgoBVaypsgq-zk8jZzNayl1TVQuencxMQxls87GUgFeUIb3bUXVcsrCwAh4zQ6QfGyuBVdAj40htvbuKMqd1~WYDICQ__&Key-Pair-Id=K3EI6M078Z3AC3","height":1380,"width":1926},"inferredFromUrl":false},"qa_pairs":{"kind":"list like","value":{"question":["What sector has the highest percentage of volunteer activities?","What is the combined percentage of volunteer activities in the Healthcare and Community Improvement sectors?","Which sectors have an equal percentage of volunteer activities?","How much greater is the percentage of volunteer activities in Education compared to Social Services?","What is the least common sector for volunteer activities?","What is the average percentage of volunteer activities across all sectors?","Which sector is represented by the green-colored wedge in the pie chart?","How does the percentage of volunteer activities in Arts and Culture compare to Environmental?","If the categories of Animal Welfare and Other were combined, what would be their total percentage?"],"explanation":["The sector with the highest percentage can be identified by locating the largest wedge. From the figure, the section that is exploded is the largest one, which represents Education.","The combined percentage can be calculated by adding the percentages of both sectors. Healthcare is 15% and Community Improvement is also 15%. Therefore, 15% + 15% = 30%.","By examining the figure, we can see that both the Animal Welfare and the Other sectors have an equal slice size, which is 5%. Similarly, Healthcare and Community Improvement both have 15%.","The percentage for Education is 25%, and for Social Services, it is 20%. The difference is calculated as 25% - 20% = 5%.","The least common sector can be identified by locating the smallest wedge in the pie chart. In the figure, the smallest sections are Animal Welfare and Other, both at 5%.","To find the average, sum all the percentages and divide by the number of sectors. The sum is 25% + 15% + 10% + 20% + 10% + 5% + 15% + 5% = 105%. There are 8 sectors, so the average is 105% / 8 = 13.125%.","By identifying the visual color correspondence, the green-colored wedge represents the Environmental sector.","Both sectors have the same percentage of 10%, as indicated by the equal size of their wedges.","Combine the percentages of Animal Welfare and Other. Both are 5%, so 5% + 5% = 10%."],"answer":["Education","30%","Animal Welfare and Other; Healthcare and Community Improvement","5%","Animal Welfare and Other","13.125%","Environmental","They are equal","10%"]},"string":"{\n \"question\": [\n \"What sector has the highest percentage of volunteer activities?\",\n \"What is the combined percentage of volunteer activities in the Healthcare and Community Improvement sectors?\",\n \"Which sectors have an equal percentage of volunteer activities?\",\n \"How much greater is the percentage of volunteer activities in Education compared to Social Services?\",\n \"What is the least common sector for volunteer activities?\",\n \"What is the average percentage of volunteer activities across all sectors?\",\n \"Which sector is represented by the green-colored wedge in the pie chart?\",\n \"How does the percentage of volunteer activities in Arts and Culture compare to Environmental?\",\n \"If the categories of Animal Welfare and Other were combined, what would be their total percentage?\"\n ],\n \"explanation\": [\n \"The sector with the highest percentage can be identified by locating the largest wedge. From the figure, the section that is exploded is the largest one, which represents Education.\",\n \"The combined percentage can be calculated by adding the percentages of both sectors. Healthcare is 15% and Community Improvement is also 15%. Therefore, 15% + 15% = 30%.\",\n \"By examining the figure, we can see that both the Animal Welfare and the Other sectors have an equal slice size, which is 5%. Similarly, Healthcare and Community Improvement both have 15%.\",\n \"The percentage for Education is 25%, and for Social Services, it is 20%. The difference is calculated as 25% - 20% = 5%.\",\n \"The least common sector can be identified by locating the smallest wedge in the pie chart. In the figure, the smallest sections are Animal Welfare and Other, both at 5%.\",\n \"To find the average, sum all the percentages and divide by the number of sectors. The sum is 25% + 15% + 10% + 20% + 10% + 5% + 15% + 5% = 105%. There are 8 sectors, so the average is 105% / 8 = 13.125%.\",\n \"By identifying the visual color correspondence, the green-colored wedge represents the Environmental sector.\",\n \"Both sectors have the same percentage of 10%, as indicated by the equal size of their wedges.\",\n \"Combine the percentages of Animal Welfare and Other. Both are 5%, so 5% + 5% = 10%.\"\n ],\n \"answer\": [\n \"Education\",\n \"30%\",\n \"Animal Welfare and Other; Healthcare and Community Improvement\",\n \"5%\",\n \"Animal Welfare and Other\",\n \"13.125%\",\n \"Environmental\",\n \"They are equal\",\n \"10%\"\n ]\n}"},"metadata":{"kind":"string","value":"{\n \"figure_type\": \"pie chart\",\n \"persona\": \"a resident of Vicksburg, Mississippi, who values community service and has a background in social work.\",\n \"topic\": \"Distribution of Volunteer Activities by Sector in Vicksburg, Mississippi\"\n}"},"data":{"kind":"string","value":"sector,percentage\nEducation,25\nHealthcare,15\nEnvironmental,10\nSocial Services,20\nArts and Culture,10\nAnimal Welfare,5\nCommunity Improvement,15\nOther,5"},"code":{"kind":"string","value":"import pandas as pd\nimport matplotlib.pyplot as plt\nfrom io import BytesIO\nfrom PIL import Image\n\ndef generate_plot(df):\n # Set the style\n plt.style.use('default')\n \n # Pie chart, where the slices will be ordered and plotted counter-clockwise:\n labels = df['sector']\n sizes = df['percentage']\n \n # Define colors for each sector\n colors = ['#ff9999','#66b3ff','#99ff99','#ffcc99','#c2c2f0','#ffb3e6', '#c4e17f', '#ffb347']\n \n # Create explode data to highlight the largest section\n explode = [0.1 if size == max(sizes) else 0 for size in sizes]\n \n fig, ax = plt.subplots(figsize=(10, 7))\n wedges, texts, autotexts = ax.pie(sizes, explode=explode, labels=labels, colors=colors, \n autopct='%1.1f%%', startangle=140, pctdistance=0.85, \n textprops=dict(color=\"black\", fontsize=12, weight='bold'))\n\n # Draw a circle at the center of pie to make it look like a donut chart\n centre_circle = plt.Circle((0,0),0.70,fc='white')\n fig.gca().add_artist(centre_circle)\n\n # Equal aspect ratio ensures that pie is drawn as a circle\n ax.axis('equal') \n \n # Title and layout adjustments\n plt.title('Distribution of Volunteer Activities by Sector\\nin Vicksburg, Mississippi', fontsize=16, weight='bold')\n plt.tight_layout()\n \n # Save the pie chart to a BytesIO and return as PIL Image\n img_bytes = BytesIO()\n plt.savefig(img_bytes, format='png', bbox_inches='tight', dpi=200)\n img_bytes.seek(0)\n return Image.open(img_bytes)"}}},{"rowIdx":30,"cells":{"id":{"kind":"string","value":"MatplotlibChartPipeline_funnel_1-305"},"image":{"kind":"image","value":{"src":"https://datasets-server.huggingface.co/assets/allenai/CoSyn-400K/--/86e46e1fd5e754d056169f0fb38f06c6997ff7de/--/chart/train/30/image/image.jpg?Expires=1741922929&Signature=Mn574xNpH1NQaz1c1TEOWbheXdS6aRqisvJg19FtroAZ95yG5xBWrTHlQcFQaCNJXcxp-8wSBOwg7PpcP5GI2fjwMzWyghH-mtFl5tMvJy~DOu-lEPE-wAtDqGdp76gLk0dwnFskeByJCQdGLiM2Qelov3JEPq1Nyu0V5YpomctgLyDJbXUnpxoUO2tP1~ehYfE0tgq1OA5dvLycc9gXuZMAP0B8mjzt7M2CNcTSnsQNHhZX2nz1xzKJfkYOVUN2azDQXS7HcOjv-LmMkbXuznFIyrdgEGtE0dzxnbO2Hon-z7wWp5~d4euieYbLfPgYTwGGntWZ7ESZuBQxN4BIaw__&Key-Pair-Id=K3EI6M078Z3AC3","height":1440,"width":1804},"inferredFromUrl":false},"qa_pairs":{"kind":"list like","value":{"question":["What's the title of the figure?","How many aircraft models are represented in the funnel chart?","Which aircraft model has the highest cumulative fuel efficiency percentage?","What is the flight altitude of the Pitts Special S-2B?","What cumulative fuel efficiency percentage does the Red Bull Air Race plane contribute?","By how much does the fuel efficiency of the Zivko Edge 540 differ from the P-51 Mustang?","Which aircraft model has the lowest cumulative fuel efficiency percentage?","How does the flight altitude of the Extra 300 compare to that of the Sukhoi Su-26?","What's the average cumulative fuel efficiency percentage of the top three aircraft models?","How many aircraft models have a flight altitude of 5000 ft or higher?"],"explanation":["The title of the figure is located at the top and is typically the most prominent text.","There are eight horizontal bars on the chart, each representing a different aircraft model.","The aircraft model at the top of the funnel chart has the highest cumulative fuel efficiency percentage.","The flight altitude is listed next to the aircraft model name on the y-axis.","The cumulative percentage for each model is shown at the end of each bar. The Red Bull Air Race plane is listed near the bottom.","Subtract the cumulative percentage of the P-51 Mustang from the Zivko Edge 540. The P-51 Mustang has approximately 51.6%, and the Zivko Edge 540 is around 86.0%.","The aircraft model at the bottom of the funnel chart has the lowest percentage.","Compare the listed flight altitudes next to the aircraft model names on the y-axis.","The top three aircraft models are Boeing Stearman, Extra 300, and Pitts Special S-2B. Their percentages are approximately 100%, 79.7%, and 67.6%, respectively. (100 + 79.7 + 67.6) / 3 = 82.43%","Examine the flight altitudes next to the aircraft model names on the y-axis and count those that are 5000 ft or higher. These are F4U Corsair (5000 ft), Sukhoi Su-26 (6000 ft), Red Bull Air Race plane (7000 ft), and Zivko Edge 540 (8000 ft)."],"answer":["Aviation Fuel Efficiency Funnel Chart","8","Boeing Stearman","4000 ft","91.1%","34.4%","F4U Corsair","3000 ft is lower than 6000 ft","82.43%","4"]},"string":"{\n \"question\": [\n \"What's the title of the figure?\",\n \"How many aircraft models are represented in the funnel chart?\",\n \"Which aircraft model has the highest cumulative fuel efficiency percentage?\",\n \"What is the flight altitude of the Pitts Special S-2B?\",\n \"What cumulative fuel efficiency percentage does the Red Bull Air Race plane contribute?\",\n \"By how much does the fuel efficiency of the Zivko Edge 540 differ from the P-51 Mustang?\",\n \"Which aircraft model has the lowest cumulative fuel efficiency percentage?\",\n \"How does the flight altitude of the Extra 300 compare to that of the Sukhoi Su-26?\",\n \"What's the average cumulative fuel efficiency percentage of the top three aircraft models?\",\n \"How many aircraft models have a flight altitude of 5000 ft or higher?\"\n ],\n \"explanation\": [\n \"The title of the figure is located at the top and is typically the most prominent text.\",\n \"There are eight horizontal bars on the chart, each representing a different aircraft model.\",\n \"The aircraft model at the top of the funnel chart has the highest cumulative fuel efficiency percentage.\",\n \"The flight altitude is listed next to the aircraft model name on the y-axis.\",\n \"The cumulative percentage for each model is shown at the end of each bar. The Red Bull Air Race plane is listed near the bottom.\",\n \"Subtract the cumulative percentage of the P-51 Mustang from the Zivko Edge 540. The P-51 Mustang has approximately 51.6%, and the Zivko Edge 540 is around 86.0%.\",\n \"The aircraft model at the bottom of the funnel chart has the lowest percentage.\",\n \"Compare the listed flight altitudes next to the aircraft model names on the y-axis.\",\n \"The top three aircraft models are Boeing Stearman, Extra 300, and Pitts Special S-2B. Their percentages are approximately 100%, 79.7%, and 67.6%, respectively. (100 + 79.7 + 67.6) / 3 = 82.43%\",\n \"Examine the flight altitudes next to the aircraft model names on the y-axis and count those that are 5000 ft or higher. These are F4U Corsair (5000 ft), Sukhoi Su-26 (6000 ft), Red Bull Air Race plane (7000 ft), and Zivko Edge 540 (8000 ft).\"\n ],\n \"answer\": [\n \"Aviation Fuel Efficiency Funnel Chart\",\n \"8\",\n \"Boeing Stearman\",\n \"4000 ft\",\n \"91.1%\",\n \"34.4%\",\n \"F4U Corsair\",\n \"3000 ft is lower than 6000 ft\",\n \"82.43%\",\n \"4\"\n ]\n}"},"metadata":{"kind":"string","value":"{\n \"figure_type\": \"Funnel Chart\",\n \"persona\": \"A fearless stunt pilot, inspired by her grandfather's wartime experiences\",\n \"topic\": \"Aviation fuel efficiency across different aircraft models and flight altitudes\"\n}"},"data":{"kind":"string","value":"Aircraft Model,Flight Altitude (ft),Fuel Efficiency (mpg)\nF4U Corsair,5000,2.5\nP-51 Mustang,10000,3.2\nBoeing Stearman,2000,8.5\nExtra 300,3000,6.8\nPitts Special S-2B,4000,5.7\nSukhoi Su-26,6000,4.9\nZivko Edge 540,8000,4.3\nRed Bull Air Race plane,7000,4.6"},"code":{"kind":"string","value":"import numpy as np\nimport matplotlib.pyplot as plt\nfrom io import BytesIO\nfrom PIL import Image\n\ndef generate_plot(df):\n plt.style.use('default')\n \n fig, ax = plt.subplots(figsize=(10, 8))\n \n # Sort data by Fuel Efficiency\n df_sorted = df.sort_values('Fuel Efficiency (mpg)', ascending=False)\n \n # Calculate cumulative percentages\n total = df_sorted['Fuel Efficiency (mpg)'].sum()\n cumulative_percentages = df_sorted['Fuel Efficiency (mpg)'].cumsum() / total * 100\n \n # Create funnel chart\n y_pos = range(len(df_sorted))\n ax.barh(y_pos, cumulative_percentages, alpha=0.8, color='skyblue', edgecolor='navy')\n \n # Customize plot\n ax.set_yticks(y_pos)\n ax.set_yticklabels(df_sorted['Aircraft Model'], fontsize=10)\n ax.set_xlabel('Cumulative Fuel Efficiency (%)', fontsize=12, fontweight='bold')\n ax.set_title('Aviation Fuel Efficiency Funnel Chart', fontsize=16, fontweight='bold', pad=20)\n \n # Add value labels\n for i, v in enumerate(cumulative_percentages):\n ax.text(v + 1, i, f'{v:.1f}%', va='center', fontsize=9, fontweight='bold')\n \n # Add flight altitude information\n for i, (model, altitude) in enumerate(zip(df_sorted['Aircraft Model'], df_sorted['Flight Altitude (ft)'])):\n ax.text(-15, i, f'{altitude} ft', va='center', ha='right', fontsize=8, color='darkgreen')\n \n # Adjust layout and style\n ax.spines['top'].set_visible(False)\n ax.spines['right'].set_visible(False)\n ax.spines['left'].set_visible(False)\n ax.set_xlim(0, 105)\n \n plt.tight_layout()\n \n # Save plot to BytesIO\n buf = BytesIO()\n plt.savefig(buf, format='png', dpi=300, bbox_inches='tight')\n buf.seek(0)\n \n # Convert to PIL Image\n img = Image.open(buf)\n \n return img"}}},{"rowIdx":31,"cells":{"id":{"kind":"string","value":"MatplotlibChartPipeline_bullet_1-162"},"image":{"kind":"image","value":{"src":"https://datasets-server.huggingface.co/assets/allenai/CoSyn-400K/--/86e46e1fd5e754d056169f0fb38f06c6997ff7de/--/chart/train/31/image/image.jpg?Expires=1741922929&Signature=JBmtarHQSYNyNehkXo7ZveE9ZdrF1PEERxWCa1rHo0MqpuIZJJogpI8tqQ9OF4lNp9hIcs1~FRoyI52AmxD85X5EC4RG75O1Y6hOqttuDJEVNqVFNxFcD47p5r86fElqXNwRyRKZp3XKOdJeFZ7nrdwBvBSec4VvzcUkUgNLnUgoCx2PhlrszwQD9RDeA2axcI9X5XnCttec8PEAgpveagvZhkEFRrMaDpF7YE1DiMbyJ-JkmlimWEvxgIb5ECHLaQwoct3dtoSqjBzZZp2h~LRbY6Fi8z6SimG0NuTYsNnh-CNdtzTwZQqaI4gblzUdROc~3a5nDC8w58-xNac~mw__&Key-Pair-Id=K3EI6M078Z3AC3","height":1440,"width":2169},"inferredFromUrl":false},"qa_pairs":{"kind":"list like","value":{"question":["What is the title of the figure?","Which country has the highest actual progress percentage?","For India, what is the difference between the target percentage and the actual percentage?","Which country has the smallest difference between the benchmark and the actual progress?","What is the benchmark percentage for Turkey?","How does Brazil's actual progress compare to Nigeria's actual progress?","Which countries have actual progress percentages below their benchmarks?","What is the mean of the target percentages for the given countries?","Which country has the lowest actual progress relative to its target?"],"explanation":["The title is usually displayed prominently at the top of the figure. In this case, it reads \"Progress towards Religious Freedom Legislation Targets by Country\".","To find this, look at the lengths of the green bars for each country. The United States has the longest green bar, indicating the highest actual progress.","To calculate this, subtract the actual percentage from the target percentage for India: 80 (target) - 45 (actual) = 35.","To find this, look for the smallest distance between the benchmark (yellow diamond) and the actual progress (green bar). The Philippines shows the smallest difference between these two.","The benchmark percentage is indicated by the yellow diamond for each country. For Turkey, this diamond is positioned at 55%.","Brazil's actual progress (green bar) is higher at 70%, whereas Nigeria's actual progress is lower at 30%.","Countries with green bars shorter than their yellow diamonds indicate actual progress below benchmarks. These countries are India, Indonesia, Nigeria, Russia, Pakistan, Turkey, and Egypt.","Add all target percentages and divide by the number of countries: (90 + 80 + 70 + 60 + 50 + 85 + 55 + 65 + 45 + 75) / 10 = 67.5.","Calculate the difference between the actual percentage and the target percentage for each country, then identify the largest negative difference. Russia has the largest negative difference, with a target of 50% and an actual progress of 20%, resulting in -30."],"answer":["Progress towards Religious Freedom Legislation Targets by Country","United States","35","Philippines","55%","Brazil's actual progress is higher than Nigeria's","India, Indonesia, Nigeria, Russia, Pakistan, Turkey, Egypt","67.5","Russia"]},"string":"{\n \"question\": [\n \"What is the title of the figure?\",\n \"Which country has the highest actual progress percentage?\",\n \"For India, what is the difference between the target percentage and the actual percentage?\",\n \"Which country has the smallest difference between the benchmark and the actual progress?\",\n \"What is the benchmark percentage for Turkey?\",\n \"How does Brazil's actual progress compare to Nigeria's actual progress?\",\n \"Which countries have actual progress percentages below their benchmarks?\",\n \"What is the mean of the target percentages for the given countries?\",\n \"Which country has the lowest actual progress relative to its target?\"\n ],\n \"explanation\": [\n \"The title is usually displayed prominently at the top of the figure. In this case, it reads \\\"Progress towards Religious Freedom Legislation Targets by Country\\\".\",\n \"To find this, look at the lengths of the green bars for each country. The United States has the longest green bar, indicating the highest actual progress.\",\n \"To calculate this, subtract the actual percentage from the target percentage for India: 80 (target) - 45 (actual) = 35.\",\n \"To find this, look for the smallest distance between the benchmark (yellow diamond) and the actual progress (green bar). The Philippines shows the smallest difference between these two.\",\n \"The benchmark percentage is indicated by the yellow diamond for each country. For Turkey, this diamond is positioned at 55%.\",\n \"Brazil's actual progress (green bar) is higher at 70%, whereas Nigeria's actual progress is lower at 30%.\",\n \"Countries with green bars shorter than their yellow diamonds indicate actual progress below benchmarks. These countries are India, Indonesia, Nigeria, Russia, Pakistan, Turkey, and Egypt.\",\n \"Add all target percentages and divide by the number of countries: (90 + 80 + 70 + 60 + 50 + 85 + 55 + 65 + 45 + 75) / 10 = 67.5.\",\n \"Calculate the difference between the actual percentage and the target percentage for each country, then identify the largest negative difference. Russia has the largest negative difference, with a target of 50% and an actual progress of 20%, resulting in -30.\"\n ],\n \"answer\": [\n \"Progress towards Religious Freedom Legislation Targets by Country\",\n \"United States\",\n \"35\",\n \"Philippines\",\n \"55%\",\n \"Brazil's actual progress is higher than Nigeria's\",\n \"India, Indonesia, Nigeria, Russia, Pakistan, Turkey, Egypt\",\n \"67.5\",\n \"Russia\"\n ]\n}"},"metadata":{"kind":"string","value":"{\n \"figure_type\": \"Bullet Chart\",\n \"persona\": \"A representative from a different religious organization who stands in solidarity with the religious leader and supports religious freedom\",\n \"topic\": \"Progress towards religious freedom legislation targets by country\"\n}"},"data":{"kind":"string","value":"Country,Target,Actual,Benchmark\nUnited States,90,75,80\nIndia,80,45,60\nIndonesia,70,55,65\nNigeria,60,30,50\nRussia,50,20,40\nBrazil,85,70,75\nPakistan,55,25,45\nTurkey,65,40,55\nEgypt,45,15,35\nPhilippines,75,60,70"},"code":{"kind":"string","value":"import numpy as np\nimport matplotlib.pyplot as plt\nfrom io import BytesIO\nfrom PIL import Image\n\ndef generate_plot(data):\n plt.style.use('default')\n fig, ax = plt.subplots(figsize=(12, 8))\n \n countries = data['Country']\n targets = data['Target']\n actuals = data['Actual']\n benchmarks = data['Benchmark']\n \n y_pos = np.arange(len(countries))\n bar_height = 0.5\n \n # Plot actuals\n ax.barh(y_pos, actuals, height=bar_height, color='#4CAF50', alpha=0.8, label='Actual')\n \n # Plot targets\n ax.barh(y_pos, targets, height=bar_height/3, color='#2196F3', alpha=0.9, label='Target')\n \n # Plot benchmarks\n ax.scatter(benchmarks, y_pos, marker='D', s=100, color='#FFC107', label='Benchmark')\n \n ax.set_yticks(y_pos)\n ax.set_yticklabels(countries, fontsize=10, fontweight='bold')\n ax.invert_yaxis()\n \n ax.set_xlabel('Progress (%)', fontsize=12, fontweight='bold')\n ax.set_title('Progress towards Religious Freedom Legislation Targets by Country', \n fontsize=16, fontweight='bold', pad=20)\n \n ax.legend(loc='lower right', fontsize=10)\n \n ax.spines['top'].set_visible(False)\n ax.spines['right'].set_visible(False)\n ax.spines['left'].set_visible(False)\n \n ax.set_axisbelow(True)\n ax.grid(axis='x', linestyle='--', alpha=0.7)\n \n ax.set_xlim(0, max(targets) * 1.1)\n \n plt.tight_layout()\n \n buf = BytesIO()\n plt.savefig(buf, format='png', dpi=300, bbox_inches='tight')\n buf.seek(0)\n img = Image.open(buf)\n \n return img"}}},{"rowIdx":32,"cells":{"id":{"kind":"string","value":"PlotlyChartPipeline_pie_claude_2-836"},"image":{"kind":"image","value":{"src":"https://datasets-server.huggingface.co/assets/allenai/CoSyn-400K/--/86e46e1fd5e754d056169f0fb38f06c6997ff7de/--/chart/train/32/image/image.jpg?Expires=1741922929&Signature=FaQtSqsOOdV4NYM6RBOEPvr9isdAd8rllQMHHRcguiTrV43d42Qcy~EN8PPRtSSUGnd-j~7JMGhZp9He4vcjJBtSzbugIuSHCFZD5Cbt0Ev92B6EYO0WTARgXOz7SsJjLl57p-rL46AJeWvxOtyIWXLUYDeEoKm1xCzAX8BbDd6ZE0SrVT4gwjXq3yrPiHH4U2TVAVpu-iPoJbOlYseP3iMss~GcV~HryrXKGAviIbBtW9hlSY7Er5I-FqmAtwa2ofu7re3AAT-TUDDUfwVOX7d-clU2DTaNbnwbJj2~hkHzOkYb6mrKGwzVA3kW255stM1ZhOvnT8nySHaxBhx-kg__&Key-Pair-Id=K3EI6M078Z3AC3","height":1600,"width":1600},"inferredFromUrl":false},"qa_pairs":{"kind":"list like","value":{"question":["What is the most common architectural style among the historic buildings?","Which architectural style has a proportion that is more than twice that of Norman?","How much more common is Gothic compared to Victorian?","If you combine the proportions of Georgian and Victorian styles, what is their total percentage?","Which architectural style is represented by the smallest segment in the pie chart?","Is the proportion of Tudor greater than the combined total of Norman and Jacobean?","What is the range of the percentages shown in the pie chart?","What proportion of historic buildings have either Gothic or Arts and Crafts styles?","By how much does the Georgian style exceed the Norman style in percentage?"],"explanation":["The figure shows the proportions of different architectural styles. Gothic has the largest segment.","Check the percentage for Norman, which is 8%. Gothic (35%) and Tudor (22%) are more than double (8% x 2 = 16%).","Comparing the percentages: Gothic is 35%, and Victorian is 12%. The difference is 35% - 12%.","Add Georgian (18%) and Victorian (12%) percentages together: 18% + 12%.","Look at the smallest segment in the pie chart, which corresponds to Arts and Crafts (2%).","Check the percentages: Tudor is 22%, Norman is 8%, and Jacobean is 3%. Combined total of Norman and Jacobean is 8% + 3% = 11%. 22% > 11%.","Take the highest percentage, Gothic (35%), and the lowest, Arts and Crafts (2%), and find the difference: 35% - 2%.","Add the percentages for Gothic (35%) and Arts and Crafts (2%) together: 35% + 2%.","Subtract the Norman percentage (8%) from the Georgian percentage (18%): 18% - 8%."],"answer":["Gothic","Gothic, Tudor","23%","30%","Arts and Crafts","Yes","33%","37%","10%"]},"string":"{\n \"question\": [\n \"What is the most common architectural style among the historic buildings?\",\n \"Which architectural style has a proportion that is more than twice that of Norman?\",\n \"How much more common is Gothic compared to Victorian?\",\n \"If you combine the proportions of Georgian and Victorian styles, what is their total percentage?\",\n \"Which architectural style is represented by the smallest segment in the pie chart?\",\n \"Is the proportion of Tudor greater than the combined total of Norman and Jacobean?\",\n \"What is the range of the percentages shown in the pie chart?\",\n \"What proportion of historic buildings have either Gothic or Arts and Crafts styles?\",\n \"By how much does the Georgian style exceed the Norman style in percentage?\"\n ],\n \"explanation\": [\n \"The figure shows the proportions of different architectural styles. Gothic has the largest segment.\",\n \"Check the percentage for Norman, which is 8%. Gothic (35%) and Tudor (22%) are more than double (8% x 2 = 16%).\",\n \"Comparing the percentages: Gothic is 35%, and Victorian is 12%. The difference is 35% - 12%.\",\n \"Add Georgian (18%) and Victorian (12%) percentages together: 18% + 12%.\",\n \"Look at the smallest segment in the pie chart, which corresponds to Arts and Crafts (2%).\",\n \"Check the percentages: Tudor is 22%, Norman is 8%, and Jacobean is 3%. Combined total of Norman and Jacobean is 8% + 3% = 11%. 22% > 11%.\",\n \"Take the highest percentage, Gothic (35%), and the lowest, Arts and Crafts (2%), and find the difference: 35% - 2%.\",\n \"Add the percentages for Gothic (35%) and Arts and Crafts (2%) together: 35% + 2%.\",\n \"Subtract the Norman percentage (8%) from the Georgian percentage (18%): 18% - 8%.\"\n ],\n \"answer\": [\n \"Gothic\",\n \"Gothic, Tudor\",\n \"23%\",\n \"30%\",\n \"Arts and Crafts\",\n \"Yes\",\n \"33%\",\n \"37%\",\n \"10%\"\n ]\n}"},"metadata":{"kind":"string","value":"{\n \"figure_type\": \"pie chart\",\n \"persona\": \"a conservative British vicar who has long served in a small rural church in the UK countryside and is also passionate about local arts and culture.\",\n \"topic\": \"Proportion of different architectural styles among historic buildings in the area\"\n}"},"data":{"kind":"string","value":"Architectural Style,Percentage\nGothic,35\nTudor,22\nGeorgian,18\nVictorian,12\nNorman,8\nJacobean,3\nArts and Crafts,2"},"code":{"kind":"string","value":"import plotly.graph_objects as go\nimport plotly.io as pio\nfrom io import BytesIO\nfrom PIL import Image\nimport numpy as np\n\ndef generate_plot(data):\n colors = ['#8B4513', '#A0522D', '#CD853F', '#DEB887', '#F4A460', '#D2691E', '#B8860B']\n \n fig = go.Figure(data=[go.Pie(\n labels=data['Architectural Style'],\n values=data['Percentage'],\n hole=.3,\n marker=dict(colors=colors, line=dict(color='#000000', width=2)),\n textfont=dict(size=14, family='Times New Roman'),\n textposition='outside',\n textinfo='label+percent',\n pull=[0.1 if x == max(data['Percentage']) else 0 for x in data['Percentage']],\n rotation=90\n )])\n\n fig.update_layout(\n title=dict(\n text='Architectural Styles of Historic Buildings',\n font=dict(size=24, family='Garamond', color='#4B0082'),\n x=0.5,\n y=0.95\n ),\n paper_bgcolor='#F5F5DC',\n plot_bgcolor='#F5F5DC',\n legend=dict(\n font=dict(size=12, family='Garamond'),\n orientation='h',\n yanchor='bottom',\n y=-0.2,\n xanchor='center',\n x=0.5\n ),\n margin=dict(t=100, b=100, l=50, r=50),\n annotations=[\n dict(\n text='Preserving our architectural heritage',\n showarrow=False,\n xref='paper',\n yref='paper',\n x=0.5,\n y=-0.15,\n font=dict(size=14, family='Garamond', color='#4B0082')\n )\n ]\n )\n\n img_bytes = BytesIO()\n pio.write_image(fig, img_bytes, format='png', width=800, height=800, scale=2)\n img_bytes.seek(0)\n return Image.open(img_bytes)"}}},{"rowIdx":33,"cells":{"id":{"kind":"string","value":"PlotlyChartPipeline_pie_claude_2-961"},"image":{"kind":"image","value":{"src":"https://datasets-server.huggingface.co/assets/allenai/CoSyn-400K/--/86e46e1fd5e754d056169f0fb38f06c6997ff7de/--/chart/train/33/image/image.jpg?Expires=1741922929&Signature=oIcMABlAEbwO0C6RuVHJwwPhQcpS9NRduJW8tTOUL539Ehg0AFfoTzd2Fsy9n3EI1S7-A77t0uV~dtcPkU0lWDY0dAWH5wtVCNN2d-cCikN1qE29TvWjAmZ2SfP8FjxbeOoc0ZG8UFt6NHBYZkds2w2a7pvv9fIcm295OmoDdip30A5ORfusruZzQiSmb9GZG22P3GsK5JmSjXDebdoAZKalHfD7PMX7hOJpmUsEVHPZm0ozXf7N9XgSgHLTKLKTAHe8SNur6nbTr73kWFZW16pzgP5hf7EW7bLrxFQgRT~0xaYjrdwKlZLQRdgXzMgmZcc4pd6OW65iFcZnL9VCnA__&Key-Pair-Id=K3EI6M078Z3AC3","height":1200,"width":1600},"inferredFromUrl":false},"qa_pairs":{"kind":"list like","value":{"question":["What percentage of time is spent on Algorithm Design?","Which phase requires more time, Code Implementation or Debugging and Testing?","What is the combined percentage of time spent on Mathematical Formulation and Performance Optimization?","Which phase is represented by the largest section of the pie chart?","How much more time is spent on Literature Review compared to Documentation?","Is the time allocated to Validation with Experimental Data higher or lower than Performance Optimization?","What is the ratio of the time spent on Code Implementation to the time spent on Debugging and Testing?","What is the percentage of time spent on phases other than Algorithm Design?","How many phases have a time allocation less than 10%?"],"explanation":["The figure indicates that Algorithm Design takes up 25% of the total time.","The figure shows that Code Implementation has 18% and Debugging and Testing has 12%. Since 18 is greater than 12, Code Implementation requires more time.","The figure shows 20% for Mathematical Formulation and 8% for Performance Optimization. Adding these together, 20 + 8 = 28%.","The figure shows various sections with their percentages clearly labeled. The largest percentage shown is for Algorithm Design at 25%.","The figure shows 15% for Literature Review and 5% for Documentation. The difference is 15% - 5% = 10%.","The figure shows 7% for Validation with Experimental Data and 8% for Performance Optimization. Since 7% is less than 8%, Validation with Experimental Data is lower.","The figure shows 18% for Code Implementation and 12% for Debugging and Testing. The ratio is 18:12. Simplifying it, we get 3:2.","The figure shows Algorithm Design at 25%. Therefore, the percentage of time spent on other phases is 100% - 25% = 75%.","The figure shows Performance Optimization at 8%, Documentation at 5%, and Validation with Experimental Data at 7%. These are the phases with less than 10%. There are 3 such phases."],"answer":["25%","Code Implementation","28%","Algorithm Design","10%","Lower","3:2","75%","3"]},"string":"{\n \"question\": [\n \"What percentage of time is spent on Algorithm Design?\",\n \"Which phase requires more time, Code Implementation or Debugging and Testing?\",\n \"What is the combined percentage of time spent on Mathematical Formulation and Performance Optimization?\",\n \"Which phase is represented by the largest section of the pie chart?\",\n \"How much more time is spent on Literature Review compared to Documentation?\",\n \"Is the time allocated to Validation with Experimental Data higher or lower than Performance Optimization?\",\n \"What is the ratio of the time spent on Code Implementation to the time spent on Debugging and Testing?\",\n \"What is the percentage of time spent on phases other than Algorithm Design?\",\n \"How many phases have a time allocation less than 10%?\"\n ],\n \"explanation\": [\n \"The figure indicates that Algorithm Design takes up 25% of the total time.\",\n \"The figure shows that Code Implementation has 18% and Debugging and Testing has 12%. Since 18 is greater than 12, Code Implementation requires more time.\",\n \"The figure shows 20% for Mathematical Formulation and 8% for Performance Optimization. Adding these together, 20 + 8 = 28%.\",\n \"The figure shows various sections with their percentages clearly labeled. The largest percentage shown is for Algorithm Design at 25%.\",\n \"The figure shows 15% for Literature Review and 5% for Documentation. The difference is 15% - 5% = 10%.\",\n \"The figure shows 7% for Validation with Experimental Data and 8% for Performance Optimization. Since 7% is less than 8%, Validation with Experimental Data is lower.\",\n \"The figure shows 18% for Code Implementation and 12% for Debugging and Testing. The ratio is 18:12. Simplifying it, we get 3:2.\",\n \"The figure shows Algorithm Design at 25%. Therefore, the percentage of time spent on other phases is 100% - 25% = 75%.\",\n \"The figure shows Performance Optimization at 8%, Documentation at 5%, and Validation with Experimental Data at 7%. These are the phases with less than 10%. There are 3 such phases.\"\n ],\n \"answer\": [\n \"25%\",\n \"Code Implementation\",\n \"28%\",\n \"Algorithm Design\",\n \"10%\",\n \"Lower\",\n \"3:2\",\n \"75%\",\n \"3\"\n ]\n}"},"metadata":{"kind":"string","value":"{\n \"figure_type\": \"pie chart\",\n \"persona\": \"A computer scientist skilled in developing simulations and algorithms for complex physical systems\",\n \"topic\": \"Breakdown of time spent on different phases of algorithm development for fluid dynamics modeling\"\n}"},"data":{"kind":"string","value":"Phase,Percentage\nAlgorithm Design,25\nLiterature Review,15\nMathematical Formulation,20\nCode Implementation,18\nDebugging and Testing,12\nPerformance Optimization,8\nDocumentation,5\nValidation with Experimental Data,7"},"code":{"kind":"string","value":"import plotly.graph_objects as go\nimport plotly.io as pio\nfrom io import BytesIO\nfrom PIL import Image\nimport numpy as np\n\ndef generate_plot(df):\n colors = ['#1f77b4', '#ff7f0e', '#2ca02c', '#d62728', '#9467bd', '#8c564b', '#e377c2', '#7f7f7f']\n \n fig = go.Figure(data=[go.Pie(\n labels=df['Phase'],\n values=df['Percentage'],\n hole=0.4,\n textinfo='label+percent',\n textfont_size=12,\n marker=dict(colors=colors, line=dict(color='#000000', width=2)),\n rotation=90\n )])\n\n fig.update_layout(\n title={\n 'text': \"Time Allocation in Fluid Dynamics Algorithm Development\",\n 'y':0.95,\n 'x':0.5,\n 'xanchor': 'center',\n 'yanchor': 'top',\n 'font': dict(size=20, color='#333333', family=\"Arial Black\")\n },\n font=dict(family=\"Arial\"),\n legend=dict(\n orientation=\"h\",\n yanchor=\"bottom\",\n y=-0.2,\n xanchor=\"center\",\n x=0.5\n ),\n margin=dict(l=20, r=20, t=80, b=20),\n paper_bgcolor='rgba(0,0,0,0)',\n plot_bgcolor='rgba(0,0,0,0)'\n )\n\n fig.update_traces(\n textposition='inside',\n textfont_size=14,\n hoverinfo='label+percent',\n pull=[0.1 if i == np.argmax(df['Percentage']) else 0 for i in range(len(df))]\n )\n\n img_bytes = BytesIO()\n pio.write_image(fig, img_bytes, format='png', width=800, height=600, scale=2)\n img_bytes.seek(0)\n return Image.open(img_bytes)"}}}],"truncated":true},"paginationData":{"pageIndex":0,"numItemsPerPage":100,"numTotalItems":116814,"offset":0,"length":100}},"jwt":"eyJhbGciOiJFZERTQSJ9.eyJyZWFkIjp0cnVlLCJwZXJtaXNzaW9ucyI6eyJyZXBvLmNvbnRlbnQucmVhZCI6dHJ1ZX0sImlhdCI6MTc0MTkxOTMyOSwic3ViIjoiL2RhdGFzZXRzL2FsbGVuYWkvQ29TeW4tNDAwSyIsImV4cCI6MTc0MTkyMjkyOSwiaXNzIjoiaHR0cHM6Ly9odWdnaW5nZmFjZS5jbyJ9.CJXE2ZYD_WxLmsPh9msSWH9NH3UeJ2T0foUIu8IIw0ppwiYcLTDqHzW6A4e0KSChP8NS9llTdjFxgd2OE5W-Aw","displayUrls":true},"dataset":"allenai/CoSyn-400K","isGated":false,"isPrivate":false,"hasParquetFormat":true,"author":{"avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/652db071b62cf1f8463221e2/CxxwFiaomTa1MCX_B7-pT.png","fullname":"Ai2","name":"allenai","type":"org","isHf":false,"isMod":false,"isEnterprise":true,"followerCount":2804},"compact":true}">

table_dataset = datasets.load_dataset("allenai/CoSyn-400K", "table", split="train")

print(table_dataset[0]["image"])
# >>> PIL.PngImagePlugin.PngImageFile image mode=RGB size=2400x1200 at 0x7F362070CEB0>

for q, a in zip(table_dataset[0]["questions"]["question"], table_dataset[0]["questions"]["answer"]):
  print(q, a)
# >>>
# What is the waist circumference range for adult females? 64-88 cm
# What is the weight range for children aged 2-12 years? 10-45 kg
# Is the BMI range for infants provided in the table? No
# Which age group has the highest resting heart rate range? Infants (0-1 year)
# What is the difference in lung capacity range between adolescents and elderly? Maximum difference: 0.5 L, Minimum difference: 1.5 L
# Do adult males have a higher blood pressure range than adolescents? Yes
# What is the average height of elderly females compared to male adolescents? Male adolescents are taller by 10 cm
# Does the table provide a consistent BMI range across all groups for females? Yes
# Which gender has a lower average hip circumference range among the elderly? Females have a lower average hip circumference

@article{yang2025scaling,
      title={Scaling Text-Rich Image Understanding via Code-Guided Synthetic Multimodal Data Generation},
      author={Yang, Yue and Patel, Ajay and Deitke, Matt and Gupta, Tanmay and Weihs, Luca and Head, Andrew and Yatskar, Mark and Callison-Burch, Chris and Krishna, Ranjay and Kembhavi, Aniruddha and others},
      journal={arXiv preprint arXiv:2502.14846},
      year={2025}
}

@article{deitke2024molmo,
  title={Molmo and pixmo: Open weights and open data for state-of-the-art multimodal models},
  author={Deitke, Matt and Clark, Christopher and Lee, Sangho and Tripathi, Rohun and Yang, Yue and Park, Jae Sung and Salehi, Mohammadreza and Muennighoff, Niklas and Lo, Kyle and Soldaini, Luca and others},
  journal={arXiv preprint arXiv:2409.17146},
  year={2024}
}