llm_provider_pipstats.py

"""
This script fetches download statistics for major LLM provider packages (OpenAI, Anthropic, Claude) from PyPI Stats API
and generates an HTML visualization showing the relative market share across different operating systems.

The visualization consists of three pie charts displaying the percentage of downloads for each package on:
- Windows
- MacOS (Darwin)
- Linux

Each chart shows:
- Package name
- Percentage of total downloads for that OS
- Absolute number of downloads
"""

import requests
import pandas as pd
import altair as alt

URL = "https://pypistats.org/api/packages/{PACKAGE}/system"
PACKAGES = ["openai", "anthropic", "google-generativeai"]
CATEGORIES = ["Windows", "Darwin", "Linux"]

def fetch_package_data(package):
    url = URL.format(PACKAGE=package)
    response = requests.get(url)
    json_data = response.json()
    return [dict(item, package=package) for item in json_data['data']]

def create_category_chart(base, category, title):
    return base.transform_filter(
        alt.datum.category == category
    ).properties(title=f'{title}')

all_data = []
for package in PACKAGES:
    all_data.extend(fetch_package_data(package))

df = pd.DataFrame(all_data)
df['date'] = pd.to_datetime(df['date'])
df = df[df['category'].isin(CATEGORIES)]

df_totals = df.groupby(['category', 'package'])['downloads'].sum().reset_index()
df_totals['percentage'] = df_totals.groupby('category')['downloads'].transform(
    lambda x: x / x.sum() * 100
)

base = alt.Chart(df_totals).mark_arc().encode(
    theta='percentage:Q',
    color='package:N',
    tooltip=['package:N', 'percentage:Q', 'downloads:Q']
).properties(
    width=300,
    height=300
)

charts = [
    create_category_chart(base, 'Windows', 'Windows'),
    create_category_chart(base, 'Darwin', 'MacOS'),
    create_category_chart(base, 'Linux', 'Linux')
]

alt.hconcat(*charts).save('llm_provider_pipstats.html')