ArviZ (pronounced "AR-vees") is a Python package for exploratory analysis of Bayesian models. Includes functions for posterior analysis, model checking, comparison and diagnostics.
The official Arviz documentation can be found here https://arviz-devs.github.io/arviz/index.html
The latest version can be installed from the master branch using pip:
pip install git+git://github.com/arviz-devs/arviz.git
Another option is to clone the repository and install using python setup.py install.
|
|
|
|
|
|
|
|
|
|
|
|
Arviz is tested on Python 3.5 and 3.6, and depends on NumPy, SciPy, xarray, and Matplotlib.
A typical development workflow is:
- Install project requirements:
pip install requirements.txt - Install additional testing requirements:
pip install requirements-dev.txt - Write helpful code and tests.
- Verify code style:
./scripts/lint.sh - Run test suite:
pytest arviz/tests - Make a pull request.
There is also a Dockerfile which helps for isolating build problems and local development.
- Install Docker for your operating system
- Clone this repo,
- Run
./scripts/start_container.sh
This should start a local docker container called arviz, as well as a Jupyter notebook server running on port 8888. The notebook should be opened in your browser automatically (you can disable this by passing --no-browser). The container will be running the code from your local copy of arviz, so you can test your changes.