# BCI Code Implementation

This repository contains the implementation of DANet for motor-imagery EEG domain adaptation using the MOABB BNCI2014‐001 dataset.

---

## 1. Environment Setup

Create and activate a new Conda environment with Python 3.10:

```bash
conda create -n bci_env python=3.10 -y
conda activate bci_env

Install the core libraries:

# NumPy
conda install numpy

• CUDA 11.8

  pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu118

• CUDA 12.6

  pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu126

• CUDA 12.8

  pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu128

# scikit-learn (for metrics, e.g. Cohen’s kappa)
conda install conda-forge::scikit-learn
# tqdm (progress bars)
conda install conda-forge::tqdm
# MOABB
pip install moabb
# Weights & Biases (optional experiment logging ,but have to install)
conda install conda-forge::wandb

Once all dependencies are installed, run:

python moabb_train.py --wandb False

• --wandb False disables Weights & Biases logging.
• Other arguments (epochs, batch size, learning rates, etc.) can be passed as flags. See python moabb_train.py --help for details.

.
├── moabb_train.py      # Training script
├── model.py            # DANet, FeatureExtractor, Critic, Classifier, etc.
├── cache/              # Cached data (.pt files)
└── README.md           # This file