How to install Py-Feat#

Basic setup#

You can either install the latest stable version from Pypi using:

pip install py-feat

Or get the latest development version by installing directly from github:

pip install git+https://github.com/cosanlab/py-feat.git

Note

Py-Feat currently supports both CPU and GPU processing on NVIDIA cards. We have experimental support for GPUs on macOS which you can try with device='auto'. However, we currently advise using the default (cpu) on macOS until PyTorch support stabilizes.

Installation issues on arm-based macOS (e.g. m1, m2 etc)#

If you’re running into issues on arm-based macOS (e.g. m1, m2) you should install pytables using one of the methods below before installing py-feat:

pip install git+https://github.com/PyTables/PyTables.git
OR
conda install pytables

Using Google Colab#

On any page in these docs, you can you can simply click on Open In Colab badges you see to open that page in Google Colab.

Note

Make sure to add and run !pip install py-feat as the first cell into the Colab notebook before running any other code!

Development setup#

If you plan on contributing to py-feat then you might want to install an editable version so that any changes you make to source files are automatically reflected:

git clone https://github.com/cosanlab/py-feat.git  
cd py-feat 
pip install -e .
pip install -r requirements-dev.txt