Neurocognitive Insights into Altered States of Consciousness

Combining Magnetoencephalography (MEG) and Machine Learning (ML) to explore the neural correlates of altered states of consciousness.

Setup

1. Create a Python virtual environment and activate it

   python3.9 -m venv .venv_neurococodelics
   source .venv_neurococodelics/bin/activate

2. Install project dependencies

   pip install --upgrade pip
   pip install -r requirements.txt

Configuration

All ML experiments are driven by YAML config files in the machine-learning/configs/ folder.
Copy an existing config and modify as needed (e.g. data paths, features, models).

Example config:

machine-learning/configs/toy_ml_config.yml

Running Analysis

Use the run_ml.py script to launch your pipeline with a chosen config:

python machine-learning/run_ml.py \
  --config machine-learning/configs/toy_ml_config.yml

This will:

• Load your data (e.g. local_data/ketamine.csv)
• Select features & targets per the config
• Run analyses (feature selection, hyperparam search, etc.)
• Save results to results/, including a global summary at results/<experiment_id>.pkl

Inspect logs and outputs in the results/ directory once complete.