Metadata-Version: 2.4
Name: chemprop-contrib
Version: 2025.11.30
Summary: Community contributions to chemprop
Author: Jackson Burns, Shih-Cheng Li
Author-email: The Chemprop Development Team <chemprop@mit.edu>
License: MIT unless otherwise noted in individual contribution files
Project-URL: source, https://github.com/chemprop/chemprop-contrib
Project-URL: PyPi, https://pypi.org/project/chemprop/
Keywords: chemistry,machine learning,property prediction,message passing neural network,graph neural network,drug discovery
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3.11
Classifier: Operating System :: OS Independent
Requires-Python: >=3.11
Description-Content-Type: text/markdown
Requires-Dist: chemprop
Provides-Extra: mcp
Requires-Dist: chemprop>=2.2.0; extra == "mcp"
Requires-Dist: fastapi>=0.116.1; extra == "mcp"
Requires-Dist: fastmcp>=2.10.6; extra == "mcp"
Requires-Dist: httpx>=0.28.1; extra == "mcp"
Requires-Dist: mcp[cli]>=1.12.2; extra == "mcp"
Requires-Dist: pandas>=2.3.1; extra == "mcp"
Requires-Dist: requests>=2.32.4; extra == "mcp"

# `chemprop-contrib`

Community-contributed extensions to the Chemprop machine learning package

# Usage

## Installation

`chemprop-contrib` can either be pip installed directly
```
pip install chemprop-contrib
```
or installed as an optional dependency when installing Chemprop
```
pip install chemprop[contrib]
```

Then, the contributed packages can be imported in Python as
```python
from chemprop_contrib import <package_name>
```

Each package includes tests, which can be used to demonstrate the use of the package.

## Available Packages

 - `moe_regressor`: Implements the ["Adaptive Mixture of Local Experts"](https://doi.org/10.1162/neco.1991.3.1.79) model for regression.

# Contributing

All community-contributed packages should fork the repository and in a branch on the fork, make a new directory under the chemprop_contrib directory with their name. Then add the following:
 - License for the code (preferably MIT)
 - All associated code files
 - Data files, if needed, and their provenance
 - Documentation (perhaps a Notebook (*.ipynb) or README.md) with instructions on how to use code contributed
 - Tests demonstrating that the code works as intended

 If your package has additional dependencies, add them to the list of optional dependencies in the pyproject.toml with the same name as your directory.

Once the package is ready, make a pull request to the main chemprop-contrib repository for a brief review.
We provide a template for these pull requests.

## Examples

For packages which _do not_ require any new dependencies, see `moe_regressor` as an example.

For packages which _do_ require additional dependencies, see `mcp` as an example.
Note that one must edit the `pyproject.toml` and _add_ these dependencies, as mentioned in [Contributing](#contributing).
