python-ohdsi

Python wrappers for (some) OHDSI tools. This project has been initiated for supporting OMOP data sources in vantage6.

Make sure you have a working R environment with the OHDSI packages installed.

Installation

Python binding

Interact with the OMOP database using a python interface.

• Install Java JDK.
• Install R: sudo apt-get install r-base (set R_HOME)
• Install OHDSI R packages

The python packages can be installed as:

pip install ohdsi-common
pip install ohdsi-sqlrender
pip install ohdsi-circe
pip install ohdsi-database-connector
pip install ohdsi-feature-extraction
pip install ohdsi-cohort-generator
pip install ohdsi-cohort-diagnostics

API service

Spin up a small webserver next to the OMOP database to allow HTTP requests to the OMOP database. You can use the prebuild image from dockerhub:

docker pull ...
docker run ...

Or you can build the image yourself:

docker build -t ohdsi-api .
docker run -p 5000:5000 ohdsi-api

Or you can run the API service directly from the source code:

pip install -r requirements.txt
python api.py

Building documentation

cd docs
export IGNORE_R_IMPORTS=True
make html

cd docs
Set-Item -Path Env:IGNORE_R_IMPORTS -Value True
make html

or you can use make livehtml to automatically rebuild the documentation when a file is changed.

You can set the IGNORE_R_IMPORTS environment variable to ignore the R imports in the documentation. This is useful when you don't have the R packages installed but want to build the documentation anyway.

Release

make set-version VERSION=x.x.x

• If you created a new package add it to the Maketools
• Create a tag: vx.x.x
• Create release on Github using the tag you just created

The pipelines should build the Python packages and release the packages to PyPI.

User Documentation

The user documentation can be found at readthedocs.