ECI1

ECI1 is a research code repository for a continuous-speech EEG project on phoneme-level neural encoding of emotional speech in normal-hearing listeners and cochlear implant listeners.

The repository is intended to track code, notebooks without outputs, documentation, and small configuration files. Raw data, processed EEG data, generated audio, model outputs, and participant-related files stay local and are excluded from Git.

Project Folders

Folder	Purpose	Git policy
1_studysetup/	Study setup files and stimulus order metadata	Track only small source metadata; generated stimuli are ignored
2_data/	Raw and processed EEG data	Ignored
3_code/	MATLAB, Python, and Jupyter analysis code	Tracked
4_presentation/	Presentation source and figures	Track source selectively; rendered large files are ignored
5_writing/	Writing drafts and manuscript notes	Prefer Markdown docs in docs/; Word/PDF files are ignored
emo_audio/	Original and generated emotional speech audio	Ignored
resources/	MATLAB project metadata	Tracked if small and useful
docs/	Project documentation and development notes	Tracked
config/	Example shared pipeline configuration	Tracked

Current Workflow

The project workflow follows the research plan:

1. Create and prepare emotional speech stimuli.
2. Run acoustic and phoneme-level analyses.
3. Preprocess continuous-speech EEG data.
4. Estimate TRF models.
5. Extract PRPs.
6. Run statistical analysis and visualization.

The currently revised code sections are:

• 3_code/1_create_stimuli
• 3_code/3_preprocessing

Later stages are still being adapted from the pilot project.

Python Environment Setup

This project uses a project-level Python virtual environment in .venv/. The environment is local to this folder and is ignored by Git.

From the project root, activate the environment:

cd C:\projects\ECI1
.\.venv\Scripts\Activate.ps1

Install or refresh the project packages from requirements.txt:

python -m pip install -r requirements.txt

After activation, run Python scripts from the project root. For example:

python 3_code\2_acoustic_analysis\extract_phoneme_acoustic_contours.py

VS Code is configured to use .venv\Scripts\python.exe for this workspace. If VS Code does not pick it up immediately, run Python: Select Interpreter and choose the interpreter inside C:\projects\ECI1\.venv.

Git Setup

Initialize local Git from the project root:

git init
git branch -M main
git status --short

Create an empty private repository named ECI1 on GitHub, then connect it:

git remote add origin <your-private-github-repo-url>
git push -u origin main

Before each commit, check what will be tracked:

git status --short
git status --ignored

Only code, documentation, small config files, and output-stripped notebooks should appear as tracked files.

Notebook Output Policy

Jupyter notebooks are tracked, but notebook outputs should not be tracked.

Install the notebook stripping tools in the Python environment you use for this project:

python -m pip install pre-commit nbstripout
pre-commit install
nbstripout --install

Then test the setup:

nbstripout --status
pre-commit run --all-files

If a notebook is large, clear output before committing.

Data Safety

Do not commit:

• Raw participant EEG files
• Processed EEG files
• Generated .mat, .set, .fdt, .pkl, or .pickle files
• Original or generated audio
• Grant applications, drafts, consent documents, or participant-identifying material
• Notebook outputs

See docs/data_management.md for details.