import pandas as pd import matplotlib.pyplot as plt import seaborn as sns import numpy as np from scipy.interpolate import make_interp_spline # Load the CSV file file_path = '/Users/admin/Downloads/satisfaction_df_3.csv' data = pd.read_csv(file_path) # Group the data by 'param' and get the 'selected_num' values grouped_data = data.groupby('param')['selected_num'].apply(list) # Create a DataFrame for boxplot boxplot_data = pd.DataFrame(dict([(k, pd.Series(v)) for k, v in grouped_data.items()])) # Calculate means for each group means = boxplot_data.mean() # Prepare data for the smooth curve x = np.arange(len(means)) x_smooth = np.linspace(x.min(), x.max(), 200) spl = make_interp_spline(x, means, k=3) # k=3 for cubic spline means_smooth = spl(x_smooth) # Plot the boxplot plt.figure(figsize=(12, 8)) sns.set(style="whitegrid") # Define custom colors box_colors = [(219/255, 237/255, 243/255) for _ in range(len(boxplot_data.columns))] # Create the boxplot with adjusted width and custom colors ax = sns.boxplot(data=boxplot_data, linewidth=2.5, palette=box_colors, width=0.2, medianprops=dict(color='black', linewidth=2.5)) # Add smooth mean line plt.plot(x_smooth, means_smooth, color='red', label='Mean', linewidth=2.5) # Add mean points plt.scatter(x, means, color='red') # Customize the plot ax.set_ylabel('Number of parameterized parking spots', fontsize=27) ax.set_xlabel('Maximum allowed number of parameterized parking spots', fontsize=27) ax.set_xticklabels([10, 25, 40, 55], fontsize=30) ax.tick_params(axis='x', labelsize=30) # Increase x-axis tick label size ax.tick_params(axis='y', labelsize=30) # Increase y-axis tick label size ax.grid(True, axis='y', linestyle='--', linewidth=0.7) # Add legend for median line and mean line handles, labels = ax.get_legend_handles_labels() # Create custom legend entries median_handle = plt.Line2D([], [], color='black', linewidth=2.5, linestyle='-', label='Median') mean_handle = plt.Line2D([], [], color='red', linewidth=2.5, linestyle='-', label='Mean') # Add the custom legend entries manually handles.extend([median_handle, mean_handle]) labels.extend(['Median', 'Mean']) # Remove duplicates while maintaining order unique_labels_handles = dict(zip(labels, handles)) ax.legend(unique_labels_handles.values(), unique_labels_handles.keys(), fontsize=30, loc='upper left') plt.tight_layout() plt.savefig('boxplot.pdf') # Display the plot plt.show()