PADO: PAthological Data Obsession

Welcome to pado :wave:, a dataset library for accessing histopathological datasets in a standardized way from Python.

pado's goal is to provide a unified way to access data from diverse datasets. Its scope is very small and the design tries to keep everything simple.

As always: If pado is not pythonic, unintuitive, slow or if its documentation is confusing, it's a bug in pado. Feel free to report any issues or feature requests in the issue tracker!

Development happens on github :octocat:

Quickstart

To quickly get a pado dataset, for testing and familiarizing with the interface you can create a fake dataset, that's also used in the internal tests.

>>> from pado.mock import mock_dataset
>>> ds = mock_dataset(None)
>>> ds
PadoDataset('memory://pado-f5869e41-5246-4378-9057-96fda1c40edf', mode='r+')

This creates a test dataset in memory with 3 images and some fake metadata

>>> len(ds)
3
>>> ds.index
(ImageId('mock_image_0.svs', site='mock'),
 ImageId('mock_image_1.svs', site='mock'),
 ImageId('mock_image_2.svs', site='mock'))
>>> ds[0].image
Image(...)
>>> ds[0].metadata
                                          A  B  C  D
ImageId('mock_image_0.svs', site='mock')  a  2  c  4

Documentation

The documentation is currently provided in this repository and has to be build via sphinx. It'll be available online soon.

To build it, in the repository root, run

python -m pip install -e ".[docs]"
cd docs
make html

Access the documentation then at docs/build/html/index.html

Development Installation

pado can be installed directly via pip:

pip install "git+https://github.com/Bayer-Group/pado@main#egg=pado[cli,create]"

or for development you can clone and install via:

git clone https://github.com/Bayer-Group/pado.git
cd pathdrive-pado
pip install -e ".[cli,create,dev]"

if you prefer conda environments:

git clone https://github.com/Bayer-Group/pado.git
cd pathdrive-pado
conda install conda-devenv
conda devenv
conda activate pado

Note that in this environment pado is already installed in development mode, so go ahead and hack.

Contributing Guidelines

• Please use numpy docstrings.
• When contributing code, please try to use Pull Requests.
• tests go hand in hand with modules on tests packages at the same level. We use pytest.
• Please install pre-commit and install the hooks by running pre-commit install in the project root folder.

You can setup your IDE to help you adhering to these guidelines.
(Santi is happy to help you setting up pycharm in 5 minutes)

Acknowledgements

Build with love by Santi Villalba and Andreas Poehlmann from the Machine Learning Research group at Bayer.

pado: copyright 2020-2022 Bayer AG