ampere-meter 0.1.1

Tooling to track and visualize engagement with the mrpowers-io organization

ampere

Tooling to track and visualize engagement with the mrpowers-io organization

---

---

Installation

ampere is available as a CLI tool and a self-hosted web application (dagster scheduled jobs, fastapi api, and plotly dash frontend)

to use the CLI:

1. clone the repository
2. uv run ampere --help to see available cli commands

[WIP] to use the web application:

disclaimer: this is a work in progress and init scripts are not yet available

1. clone the repository
2. configure required environment variables in .env - refer to .env.example
3. run uv sync --all-extras to build the full .venv
4. docker compose up --build to start the api, backend, and frontend services

---

API

documentation available at https://api-ampere.jeffbrennan.dev/docs

currently supports three endpoints:

1. /repos - list of current repos
2. /feed - recent github activity (stars, issues, prs, etc.)
3. /downloads - download statistics (hourly, daily, weekly, monthly)

---

Metrics

Data

Sources

GitHub REST API

updated 4 times a day

• repos
• stars
• issues
• commits
• forks
• releases
• pull requests

Bigquery PyPi Downloads

updated daily

• python release downloads grouped by package version, python version, operating system, cloud platform

additionally gets download statistics for core project dependencies

• pyspark
• deltalake

---

Pages

home

Weekly github metrics

• stars
• issues
• commits

downloads

Weekly download statistics. Currently only includes python release downloads

• overall
• package version
• python version
• cloud platform

feed

scrolling, color-coded table of recent organization activity

• stars
• issues
• commits
• forks
• pull requests

issues

summary and details of current issues

• summary: open issues, median issue age, issues closed this month
• details: repo, author title, body, date, days old, comments

network stargazers

each star is a node, and people who have starred multiple repositories are connected to each other density of lines indicates common groups of stars edges of each cluster indicate users who have only starred one repository

network followers

each user is a node, internal (another organization user) followers and following are connected includes any user with a record in one of these tables:

• stg_stargazers
• stg_forks
• stg_commits
• stg_issues
• stg_pull_requests

blue edges are a mutual connection, grey edges are one-way color of node indicates number of followers

percentage of internal followers and following computed against user total followers and following

status

data quality page indicating the last time each data source was updated and the number of records in each table

about

repo summaries and links

---

Tech Stack

This project is build with an open source (d)ata stack:

• Dagster: data orchestration
• Dash: visualization
• Delta: raw data storage
• dbt: data transformation
• Docker: containerization
• DuckDB: database

---

TODOs

• dark mode
• page load speed
• mobile improvements
• create API
• create init scripts for first time setup
• add tests for API, CLI, and web application
• and github actions to run tests, linting checks
• publish to pypi
• add ampere to mrpowers-io organization