FAME3R: a re-implementation of the FAME3 model.

Installation

1. Create a conda environment with the required python version:

conda create --name fame3r-env python=3.10

2. Activate the environment:

conda activate fame3r-env

3. Install package:

pip install fame3r

Usage

Determining the optimal hyperparameters via k-fold cross-validation

fame3r-cv-hp-search -i INPUT_FILE -o OUTPUT_FOLDER -r RADIUS[OPTIONAL, DEFAULT=5] -n NUMFOLDS[OPTIONAL, DEFAULT=10]

Training a model

fame3r-train -i INPUT_FILE -o OUTPUT_FOLDER -r RADIUS[OPTIONAL, DEFAULT=5]

Applying a trained model on some (labeled) test data

fame3r-test -i INPUT_FILE -m MODEL_FILE -o OUTPUT_FOLDER -r RADIUS[OPTIONAL, DEFAULT=5] -t THRESHOLD[OPTIONAL, DEFAULT=0.2]

Computing the SoMs of some unlabeled data

fame3r-infer -i INPUT_FILE -m MODEL_FILE -o OUTPUT_FOLDER -r RADIUS[OPTIONAL, DEFAULT=5] -t THRESHOLD[OPTIONAL, DEFAULT=0.2]