dbgjerez / ml-github

A Github's repositories analysis

Github provides us with an API to extract the last 14 days' stats over our repositories.

This project aims to amplify and analyse them.

Clean data

This repository contains my stats into data folder. You can remove it because you will generate your own data.

Get the data

To get the data you should use the app.py file.

The best way to use it is using the virtual environment.

Configure

The application needs a GitHub token that grants privileges read-only over your public repositories.

The application reads the token from .env file. You have to create the file .env with the following data:

GITHUB_TOKEN={your-token}

Execute the code

Once you have created the .env file with the GITHUB_TOKEN variable, we can run the code.

The best way is using the virtual environment. If you don´t know how, I explaint it here

Execute the Python app:

❯ python app.py

The application will create a structure with one .csv per day:

❯ ls -lchR data
data:
total 4,0K
drwxr-xr-x 4 db db 4,0K jul 12 19:03 2022

data/2022:
total 8,0K
drwxr-xr-x 2 db db 4,0K jul 12 19:03 06
drwxr-xr-x 2 db db 4,0K jul 13 06:34 07

data/2022/06:
total 12K
-rw-r--r-- 1 db db 164 jul 12 19:03 20220628.csv
-rw-r--r-- 1 db db 192 jul 13 12:22 20220629.csv
-rw-r--r-- 1 db db 272 jul 13 12:22 20220630.csv

data/2022/07:
total 52K
-rw-r--r-- 1 db db 257 jul 13 12:22 20220701.csv
-rw-r--r-- 1 db db 195 jul 13 12:22 20220702.csv
-rw-r--r-- 1 db db 237 jul 13 12:22 20220703.csv
-rw-r--r-- 1 db db 497 jul 13 12:22 20220704.csv
-rw-r--r-- 1 db db 493 jul 13 12:22 20220705.csv
-rw-r--r-- 1 db db 559 jul 13 12:22 20220706.csv
-rw-r--r-- 1 db db 614 jul 13 12:22 20220707.csv
-rw-r--r-- 1 db db 433 jul 13 12:22 20220708.csv
-rw-r--r-- 1 db db 238 jul 13 12:22 20220709.csv
-rw-r--r-- 1 db db 304 jul 13 12:22 20220710.csv
-rw-r--r-- 1 db db 402 jul 13 12:22 20220711.csv
-rw-r--r-- 1 db db 638 jul 13 12:22 20220712.csv
-rw-r--r-- 1 db db 357 jul 13 12:22 20220713.csv

Makefile

In addition, a Makefile is provided, so you can use it:

make update

Analysis the data

The file repos_analysis.ipynb execute a light analisys over the data.

To update it, you have to use wherever editor that you prefer. In my case, I use junyper-notebook in local environments.

❯ jupyter nbconvert --execute --to notebook --inplace repos_analysis.ipynb

References

• https://github.com/PyGithub/PyGithub
• https://blog.dbgjerez.es/posts/python-virtualenv/