Select Page

Statistical analysis and HTML report

Task

What’s the association between median household income and obesity? Please generate a HTML report to include the regression analysis, r-square, p-value, and be sure to include a one paragraph report to interpret the findings.

Dataset

PA County shapefile

Result

Generated HTML report: income_obesity_regression_report

Generated Python Code

  1. # The task involves analyzing the relationship between median household income and obesity rates in Pennsylvania counties using a regression analysis.
  2. # The code should perform the analysis and generate a report with a scatter plot including a regression line.
  3. # The plot and report will be saved in the specified directory.
  4. # There should be no initialization code for QGIS in the script.
  5. # The Python code makes use of `pandas`, `seaborn`, and `statsmodels`, while ensuring correct file handling.
  7. import os
  8. import pandas as pd
  9. import seaborn as sns
  10. import statsmodels.api as sm
  11. from qgis.core import QgsVectorLayer
  13. def analyze_income_obesity():
  14.     # Path to the shapefile
  15.     shapefile_path = ‘C:/Users/17036/Downloads/DEMO/PennsylvaniaCounties/PennsylvaniaCounties.shp’
  16.     layer = QgsVectorLayer(shapefile_path, “Pennsylvania Counties”, “ogr”)
  18.     # Extract relevant fields
  19.     features = layer.getFeatures()
  20.     data = [{‘Median_hou’: f[‘Median_hou’], ‘Obesity’: f[‘Obesity’]} for f in features]
  22.     # DataFrame for analysis
  23.     df = pd.DataFrame(data)
  25.     # Convert fields to numeric
  26.     df[‘Median_hou’] = pd.to_numeric(df[‘Median_hou’], errors=‘coerce’)
  27.     df[‘Obesity’] = pd.to_numeric(df[‘Obesity’], errors=‘coerce’)
  28.     df.dropna(inplace=True)
  30.     # Regression Analysis
  31.     X = df[‘Median_hou’]
  32.     y = df[‘Obesity’]
  33.     X = sm.add_constant(X)
  34.     model = sm.OLS(y, X).fit()
  36.     # Scatter plot with regression line
  37.     sns.set(style=“darkgrid”)
  38.     plot = sns.lmplot(x=‘Median_hou’, y=‘Obesity’, data=df, height=6, aspect=1.5)
  40.     # Save the plot to HTML
  41.     output_dir = ‘C:/Users/17036/AppData/Roaming/QGIS/QGIS3/profiles/default/python/plugins/SpatialAnalysisAgent-master/Default_workspace’
  42.     plot_filepath = os.path.join(output_dir, “income_obesity_regression.png”)
  43.     plot.savefig(plot_filepath)
  45.     # Print the plot file path
  46.     print(plot_filepath)
  48.     # Generate HTML report
  49.     report_html = f“””
  50.    <html>
  51.    <head><h1>Regression Analysis Report</h1></head>
  52.    <body>
  53.    <p>Below is the summary of the regression analysis conducted on the association between Median Household Income and Obesity Rates:</p>
  54.    <h2>Regression Analysis Summary</h2>
  55.    <pre>{model.summary().as_text()}</pre>
  56.    <p>Our analysis indicates a relationship between median household income and obesity rates at the county level in Pennsylvania. Although the R-squared value suggests the variability in obesity that is explained by income is not very strong, it does indicate a trend that necessities further investigation.</p>
  57.    <h2>Regression Plot</h2>
  58.    <img src=”{plot_filepath}” alt=”Regression Plot”/>
  59.    </body>
  60.    </html>
  61.    “””
  63.     report_path = os.path.join(output_dir, “income_obesity_regression_report.html”)
  64.     with open(report_path, “w”) as file:
  65.         file.write(report_html)
  67.     # Print the report file path
  68.     print(report_path)
  70. # Execute the function
  71. analyze_income_obesity()