Core MSK Clinical Data Mining group code.

New home for cdm-utilities with more flexible package layout/structure

Requirements

For now the library requires python version at least 3.9. If your install gives an error about not being able to find compatible version of a library, check

python --version

If you're using an older python version, you'll need to upgrade.

Using/installing this repo.

Direct installation into your virtual environment

The simplest way to use this repo is to set up a conda env or environment of your choice and then, after git clone-ing this directory, simply issue

make install

or

pip install .

Building a Conda environment from a environment.yml file

First, install miniconda if you haven't done so already.

A virtual environment can be defined and created through an environment.yml file. For example, the Conda environment conda-env-cdm can be created with this snippet included in environment.yml

name: conda-env-cdm
channels:
  - pytorch
  - nvidia
  - conda-forge
dependencies:
   - python == 3.10
   - pandas
   - pyyaml
   - minio
   - requests
   - ipykernel
   - conda-build
   - pip
   - pip: 
       - git+https://github.com/clinical-data-mining/msk_cdm.git

Then, run

conda env create -f environment.yml

Contributing to this repo

If you're contributing to the code please run make install_precommit_hooks from the root of the repository to install the pre-commit hooks. They will probably require you to git add a second time after trying the first round of git commit (one of the hooks is the Black linter, which modifies the source file, so you need to try to git commit twice if it does).

You can run unit tests by issuing make test from the root dir.

If you're updating this repo, please make a Pull Request via a git branch. If the change is to the core functionality (i.e. modifies the core code in this repo possibly used by other projects/peple) please request a code review.

Deploying the Documentation Page

How to run the documentation page locally

Run:

mkdocs serve --dev-addr <Default 127.0.0.1:8000>

How to deploy the documentation

Run:

mkdocs gh-deploy