- Classification folder: contains a Jupyter Notebook for the classifier evaluation
- Data folder: contains scripts and Jupyter Notebooks for preprocessing, feature extraction, and data visualization. It also contains an Old folder containing notebooks from early on in our project when we were exploring other datasets
- Deliverables folder: contains some of our visualizations, in the form of image files, HTML files, or Jupyter Notebooks
- docs folder: contains this document, some interactive figures, DREAMER_info.pdf, a document with more information on the dataset, as well as PotentialResources.md, a Markdown file we used to document links that could help with advancing the project
- Greg_tutorial folder: contains Greg_tutorial.ipynb and Greg_tutorial.py, which are respectively a demo notebook and the script built based on this notebook that Greg Kiar used during his talk on scripting in Python (May 29th 2020 course).
- images folder: contains some images used in the GitPitch presentation
- DREAMER_main.py: main script for preprocessing, feature extraction, and classifier evaluation
- LICENSE: Creative Commons CC0 1.0 Universal license
- PITCHME.md: Markdown source file for the GitPitch presentation
- run.sh: Bash script for running the preprocessing, feature extraction, and classifier evaluation
- requirements.txt: Lists the packages in the virtual environment with which all of the notebooks and scripts were run (has some unnecessary packages)
-
Request access to the DREAMER dataset on Zenodo.
-
Clone this repository
This can be done with:
git clone https://github.com/brainhack-school2020/Biosignal-Emotions-BHS-2020
Then change the working directory to Biosignal-Emotions-BHS-2020
.
- Install the required dependencies (it is recommended that you create and activate a virtual environment beforehand)
The scripts and notebooks were run with Python 3.7.6
. To download the packages from the requirements.txt
file, you can run:
pip install -r documents/requirements.txt
-
Move the
DREAMER.mat
file downloaded from Zenodo toBiosignal-Emotions-BHS-2020/Data
. -
For preprocessing, feature extraction, and classification with only the EEG data, only the ECG data, and both the EEG and ECG data, you can run:
bash run.sh
- For the notebooks, you can run
jupyter notebook
and then the path to the notebook, e.g.:
jupyter notebook Deliverables/Week3_Emot_Plot_Danielle.ipynb
Alternatively, you can open them in binder in your browser:
https://notebooks.gesis.org/binder/v2/gh/brainhack-school2020/Biosignal-Emotions-BHS-2020/85279820daf948d114d780e39d609c4f704f8cb1