github2pandas supports the aggregation of project activities in a GitHub repository and makes them available in pandas dataframes


Keywords
git, github, collaborative, code, development, mining, git-miner, git-mining-tool, learning-analytics, python
License
BSD-3-Clause-Attribution
Install
pip install github2pandas==2.0.2

Documentation

Transform GitHub Activities to Pandas Dataframes

General information

This package is being developed by the participating partners (TU Bergakademie Freiberg, OVGU Magdeburg and HU Berlin) as part of the DiP-iT project Website.

The package implements Python functions for

  • aggregating and preprocessing GitHub activities (Commits, Actions, Issues, Pull-Requests) and
  • generating project progress summaries according to different metrics (ratio of changed lines, ratio of aggregated Levenshtein distances e.g.).

github2pandas stores the collected information in a collection of pandas DataFrames starting from a user defined root folder. The structure beyond that (file names, folder names) is defined as a member variable in the corresponding classes and can be overwritten. The default configuration results in the following file structure.

|-- My_Github_Repository_0               <- Repository name
|   |- Repo.json                         <- Json file containing user and repo name
|   |- Repository
|   |   |- Repository.p  
|   |- Issues
|   |   |- pdIssuesComments.p
|   |   |- pdIssuesEvents.p
|   |   |- pdIssues.p
|   |   |- pdIssuesReactions.p
|   |- PullRequests
|   |   |- pdPullRequestsComments.p
|   |   |- pdPullRequestsCommits.p
|   |   |- pdPullRequestsEvents.p
|   |   |- pdPullRequests.p
|   |   |- pdPullRequestsReactions.p
|   |   |- pdPullRequestsReviews.p
|   |- Users.p
|   |- Versions
|   |   |- pdCommits.p
|   |   |- pdEdits.p
|   |   |- pdBranches.p
|   |   |- pVersions.db
|   |   |- repo                         <- Repository clone
|   |   |   |- ..
|   |- Workflows
|       |- pdWorkflows.p
|-- My_Github_Repository_1
...

The internal structure and relations of the data frames are included in the project's wiki.

Installation

github2pandas is available on pypi. Use pip to install the package.

global

On Linux:

sudo pip3 install github2pandas 
sudo pip install github2pandas

On Windows as admin or for one user:

pip install github2pandas
pip install --user github2pandas 

in virtual environment:

pipenv install github2pandas

Usage

GitHub token is required for use, which is used for authentication. The website describes how you can generate this for your GitHub account. Customise the username and project name and explore any public or private repository you have access to with your account!

Access token is to define in .env oder .py (.ipynb) file. The default value of python.envFile setting is ${workspaceFolder}/.env

TOKEN="example_token"

An short example of a python script:

import os
from pathlib import Path
# github2pandas imports
from github2pandas.core import Core
from github2pandas.github2pandas import GitHub2Pandas

git_repo_name = "github2pandas"
git_repo_owner = "TUBAF-IFI-DiPiT"
    
data_root_dir = Path("data")
data_root_dir.mkdir(parents=True, exist_ok=True)
github_token = os.environ['TOKEN']

github2pandas = GitHub2Pandas(github_token,data_root_dir)
repo = github2pandas.get_repo(git_repo_owner, git_repo_name)
# extract complete repository
github2pandas.generate_pandas_tables(repo)

# exports pandas files to one excel file
GitHub2Pandas.save_tables_to_excel(Path(data_root_dir,git_repo_owner,git_repo_name))

Notebook examples

Currently not updated for github2pandas version 2.0.0!! The corresponding github2pandas_notebooks repository illustrates the usage with examplary investigations.

The documentation of the module is available at https://github2pandas.readthedocs.io/.

Working with pipenv

Process Command
Installation pipenv install --dev
Run specific script pipenv run python file.py
Run all Tests pipenv run python -m unittest
Run all tests in a specific folder pipenv run python -m unittest discover -s 'tests'
Run all tests with specific filename pipenv run python -m unittest discover -p 'test_*.py'
Start Jupyter server in virtual environment pipenv run jupyter notebook

For Contributors

Naming conventions: https://namingconvention.org/python/