generate_visualizations.py

import os
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns

# Create a directory for visualizations
visualization_dir = 'visualizations'
if not os.path.exists(visualization_dir):
    os.makedirs(visualization_dir)

# Load combined data
combined_data = pd.read_csv('combined_trends_data.csv', index_col=0, parse_dates=True)

# Define key global events related to climate change
global_events = {
    pd.Timestamp('2009-12-01'): 'COP15 Copenhagen',
    pd.Timestamp('2015-12-01'): 'Paris Agreement',
    pd.Timestamp('2021-11-01'): 'COP26 Glasgow'
}

# Line Graphs
def plot_line_graphs():
    for city in combined_data['City'].unique():
        city_data = combined_data[combined_data['City'] == city]
        plt.figure(figsize=(12, 6))
        for keyword in city_data.columns:
            if keyword == 'City':
                continue
            plt.plot(city_data.index, city_data[keyword], label=keyword)
        
        plt.title(f'Temporal Trends in {city}')
        plt.xlabel('Year')
        plt.ylabel('Search Interest')
        plt.legend()
        plt.grid(True)
        
        # Annotate global events
        for date, event in global_events.items():
            plt.axvline(x=date, color='gray', linestyle='--', linewidth=0.8)
            plt.text(date, plt.ylim()[1] * 0.9, event, rotation=45, fontsize=8, color='gray')
        
        plt.tight_layout()
        plt.savefig(f'{visualization_dir}/{city.lower().replace(" ", "_")}_line_graph.png')
        plt.close()

# Heatmaps
def plot_heatmaps():
    for keyword in combined_data.columns:
        if keyword == 'City':
            continue
        plt.figure(figsize=(10, 6))
        pivot_table = combined_data.pivot_table(index=combined_data.index, columns='City', values=keyword)
        sns.heatmap(pivot_table, cmap='coolwarm', cbar_kws={'label': 'Search Interest'})
        plt.title(f'Heatmap for {keyword}')
        plt.xlabel('City')
        plt.ylabel('Time')
        plt.tight_layout()
        plt.savefig(f'{visualization_dir}/{keyword}_heatmap.png')
        plt.close()

# KDE Plots
def plot_kde_plots():
    for keyword in combined_data.columns:
        if keyword == 'City':
            continue
        plt.figure(figsize=(10, 6))
        for city in combined_data['City'].unique():
            city_data = combined_data[combined_data['City'] == city][keyword]
            if city_data.var() > 0:  # Check for variance
                sns.kdeplot(city_data, label=city)
        plt.title(f'KDE Plot for {keyword}')
        plt.xlabel('Search Interest')
        plt.ylabel('Density')
        plt.legend()
        plt.tight_layout()
        plt.savefig(f'{visualization_dir}/{keyword}_kde.png')
        plt.close()

# Generate all visualizations
plot_line_graphs()
plot_heatmaps()
plot_kde_plots()

print("Visualizations complete. Check the 'visualizations' directory for results.")