PyTEAP

PyTEAP is a Python implementation of Toolbox for Emotion Analysis using Physiological signals (TEAP).

This package intends to reimplement TEAP, originally written in MATLAB, in Python, to enable interoperation with other Python packages.

Installation

To use PyTEAP, you can either clone this repository:

$ git clone https://github.com/cheulyop/PyTEAP.git
$ cd PyTEAP

or install it via pip:

$ pip install PyTEAP

Baseline classification on DEAP

As the primary goal of this package is to provide a toolbox for processing physiological signals for emotion analysis, this package includes a script to perform simple baseline classification on DEAP dataset.

There are two ways you can do baseline classification on DEAP with PyTEAP:

1. Clone this repository and run the script for baseline classification.

$ python baseline.py --root '/path/to/deap_root'

Running baseline.py will load raw datafiles from the root directory, preprocess features and target labels, perform baseline classification with four simple classifiers: 1) Gaussian Naive Bayes, 2) random voting assuming a uniform distribution between classes, 3) majority voting, 4) class ratio voting, and finally print a table showing performance of each classifier, measured with accuracy score, balanced accuracy score, and F1-score.

2. Or, install PyTEAP via pip as shown above, and use its modules and functions as you want.

The below table shows the results of baseline classification with seed=0.

* You must have access to DEAP dataset, in particular a preprocessed data in Python format to perform above baseline classification. Please contact DEAP maintainers if you need access to the dataset.