ml4pd 2022.11.9

ML4PD - an open-source libray for building Aspen-like process models via machine learning.

ml4pd

Process design with Machine Learning.

How to Install for Users

pip install ml4pd git+https://github.com/NREL/ml4pd_models.git@v1

How to Set Up for Development

git clone https://github.com/NREL/ml4pd.git
git clone https://github.com/NREL/ml4pd_models.git
mamba env create -f ml4pd/environment.yml
conda develop ml4pd
conda develop ml4pd_models

Additional GitHub repositories for docs, tests and training

• ml4pd_utils: code base for generating & preparing data for training.
• autoaspen: database for data obtained by aspen & python.

To debug docs

• Use mkdocs serve within ml4pd directory, then go to localhost:8000.
• generate_site.py gets docstrings (written in makrdown) from classes and put them in the right directory.
• To make changes to notebooks, add ml4pd to path with sys.path.append() or conda develop.

Relationship with ml4pd_models

To minimize manual work, ml4pd dependends on a specific ml4pd_models github branch. When changes are made to either ml4pd or ml4pd_models that will break compatibility, create new branch in ml4pd_models, and apply change to workflow .yml files, this README and index.md in docs.