Lightweight utilities for processing free-association responses.

Features

• Clean short text responses
• Average and manage embeddings
• Grouping / hierarchical clustering
• Projections (classical MDS, UMAP, PaCMAP)
• Label-frequency matrices and a small cross-validated OLS report

Quick start

1. Create and activate a virtual environment.

2. Install requirements:

pip install freeassoc

Development & tests

• The project uses a src/ layout. When running tests from the repo root, make the package importable with PYTHONPATH:

PYTHONPATH=src python -m pytest -q

Examples

• See examples/simple_df_example.py for a quick runnable demonstration of embedding, caching, and label inference.

Status

• Work in progress. Use at your own risk;

Compatibility

• Tested with Python 3.10.

Attribution

• This repository is a Python adaptation of the R project associatoR (https://github.com/samuelae/associatoR) originally created by Dirk Wulff.

Author

• Ali Alhosseini