JLab Data Science Toolkit

Composible elements for data science workflows

Software Requirement

• Python 3.10
• Python packages are defined in the yaml and setup.py
• This document assumes you are running at the top directory

Installing

• Pull code from repo

git clone https://github.com/JeffersonLab/jlab_datascience_core.git

• Create default conda environment setup (for mac M1):

conda env create --file tf-metal-arm64.yaml (only once)
conda activate tf-metal-evms-ultra (required every time you use the package)

• Install repo package in environment:

cd jlab_datascience_core
pip install -e . --user

Directory Organization

├── setup.py
├── datascience_toolkit
    ├── core                              : folder containing base classes
    ├── dataprep                          : a folder with code to read and prep data
    ├── models                            : a folder with code to for models
    ├── cfg                               : a folder contains configuration
    ├── utils                             : a folder contains utilities