Parse BCD data to get browser release data

generate.py

import json
from datetime import datetime
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt

# Function to convert a string to a datetime object
def convert_string_to_date(date_string):
    return datetime.strptime(date_string, "%Y-%m-%d")

# Function to convert version numbers to integers where possible
def convert_version_to_int(version):
    try:
        return int(version)
    except ValueError:
        try:
            return int(version.split('.')[0])
        except ValueError:
            return version

# Load JSON file and process the data
with open('path_to_your_data_file.json', 'r') as f:
    data = json.load(f)

# Extract 'browsers' data
browsers_data = data['browsers']

# Initialize an empty dictionary to store the release dates
release_dates = {}

# Loop over each browser
for browser, details in browsers_data.items():
    # Get the 'releases' field which contains version information
    releases = details.get('releases', {})
    
    # For each version, get the release date
    for version, version_details in releases.items():
        release_date = version_details.get('release_date')
        if release_date:
            # Convert the version to an integer where possible
            version = convert_version_to_int(version)
            # Convert the release date to a datetime object
            release_date = convert_string_to_date(release_date)
            # Add the release date to the dictionary
            if browser not in release_dates:
                release_dates[browser] = {}
            release_dates[browser][version] = release_date

# Convert the dictionary to a pandas DataFrame
df = pd.concat({k: pd.Series(v) for k, v in release_dates.items()}).reset_index()
df.columns = ['Browser', 'Version', 'Release Date']

# Calculate the time difference between each release for each browser
df['Time Difference'] = df.groupby('Browser')['Release Date'].diff()
df['Time Difference'] = df['Time Difference'].dt.days

# Filter out rows where the time difference is more than 3 standard deviations from the mean
df_no_outliers = df[np.abs(df['Time Difference'] - df['Time Difference'].mean()) <= (3 * df['Time Difference'].std())]

# Extract the year from the release date
df_no_outliers['Year'] = df_no_outliers['Release Date'].dt.year

# Filter out rows where the year is before 2008
df_no_outliers = df_no_outliers[df_no_outliers['Year'] >= 2008]

# Calculate the average time difference for each browser for each year
average_time_difference_yearly = df_no_outliers.groupby(['Browser', 'Year'])['Time Difference'].mean().reset_index()

# Pivot the data for plotting
pivot_df = average_time_difference_yearly.pivot(index='Year', columns='Browser', values='Time Difference')

# Define a color palette with more distinct colors
color_palette = ['b', 'g', 'r', 'c', 'm', 'y', 'k', 
                 '#FF5733', '#900C3F', '#DAF7A6', '#581845', '#C70039', '#FFC300', '#273746', '#641E16']

# Plot the data with the new color palette
pivot_df.plot(kind='line', marker='o', figsize=(12, 8), color=color_palette)
plt.ylabel('Average Time Between Releases (Days)')
plt.title('Average Time Between Browser Version Releases Over Time (2008 and Later, Outliers Removed)')
plt.grid(True)
plt.legend(bbox_to_anchor=(1.05, 1), loc='upper left')
plt.show()