Automated exploratory data analysis for DagsHub repositories.
RepoViz supports Python 3.10 - 3.12.
First, create a virtual environment and install the dependencies from the pyproject.toml file using your favorite package management tools.
For example, you can use pdm and pyenv as follows:
cd path/to/repo-viz
pyenv install 3.12.4
pdm use 3.12.4
pdm install
Second, use pip to install the DagsHub client in the virtual environment. This is necessary due to a dependency version conflict with ydata-profiling.
For example:
cd path/to/repo-viz
eval $(pdm venv activate)
python -m pip install --upgrade dagshub
First, start the D-Tale server in your terminal:
python ./dtale_app.py
Then, start the RepoViz server in your terminal:
python ./app.py
Finally, open your browser and navigate to http://localhost:8051/.
When the app starts you'll be prompted to provide an access token.
Once you provide the access token, you'll be able to select a repository and a data source to visualize.
The app uses D-Tale, Sweetviz, and YData Profiling to generate the visualizations.