nuMoM2b Preprocessing — Precision Health Initiative (φ)
A module for creating reproducible partitions of the nuMoM2b data set based on configuration files.
- Maintained by Alexander L. Hayes
- Documentation on https://doc.numom2b.org/en/latest/
Getting Started
Data should not be stored here.
Familiarity with Python, UNIX systems, and Git would be helpful.
Getting Organized
The running assumption is that Data/
contains the .csv files representing
the data set, and this nuMoM2b_preprocessing/
repository contains the code
and documentation.
Each might be stored as follows (though this may be tweaked through config files):
Precision-Health-Initiative/
├── Data/
│ ├── pregnancy_outcomes.csv
│ ├── Screening.csv
│ └── Visit1.csv
└── nuMoM2b_preprocessing/
└── README.md
Getting Running
Clone the repository from GitHub:
git clone git@github.com:hayesall/nuMoM2b_preprocessing.git
If you're using Anaconda, this would be a good time to create an environment:
conda create -n PHI python=3.7
... and install dependencies.
pip install -r phi/requirements.txt
nuMoM2b_preprocessing
Using -
As a Package
Installing using the setup script adds a
numom2b_preprocessing
package.python setup.py install
... which can then be imported
import numom2b_preprocessing _params = numom2b_preprocessing.parameters(config="phi_config.json") numom2b_preprocessing.run(_params)
-
As a Commandline (CLI) Tool
Installing using
setup.py
adds an entry point to your defaultbin/
directory.python setup.py install
This adds a
numom2b_preprocessing
tool which can be invoked as follows:numom2b_preprocessing --help
-
As a Submodule
nuMoM2b_preprocessing
acts as the first step in building classification pipelines. To incorporate it in later learning steps, it may be useful to have access to the methods as a submodule. For general documentation on using submodules, refer to the Git Manual.git submodule add git@github.com:hayesall/nuMoM2b_preprocessing.git
The modules in the package can then be accessed as normal.
import nuMoM2b_preprocessing.numom2b_preprocessing _params = numom2b_preprocessing.parameters(config="phi_config.json") numom2b_preprocessing.run(_params)
nuMoM2b_preprocessing Documentation
Documentation is not currently hosted at https://doc.numom2b.org/en/latest/, but local copies may be built using Sphinx.
A separate requirements file is in the documentation/
directory.
pip install -r documentation/requirements.txt
A make.bat
and Makefile
are included:
cd documentation
make html
If the build is successful, a copy will reside in a new build/html/
directory.
open build/html/index.html # (macOS)
xdg-open build/html/index.html # (Linux — "should" open with default browser)