Examples

amica provides examples covering the most common workflows, from fitting ICA on EEG recordings to using the core Python API directly.

Getting Started

If you are new to amica, we recommend following the examples in this order:

1. MNE-Python integration – fit AMICA on an EEG recording and obtain a standard mne.preprocessing.ICA object.

2. Native amica API – fit AMICA directly on a NumPy array using the core Amica class.

3. Validation example – reproduce the numerical validation experiments on the MNE sample dataset.

---

Example Gallery

The examples below are generated automatically from the scripts in the examples/ directory.

• Examples
  • Run on your own data
  • Installation

---

Running Examples Locally

Clone the repository and install the optional dependencies:

git clone https://github.com/snesmaeili/amica.git
cd amica

pip install -e ".[all]"

or with uv:

uv pip install -e ".[all]"

You can then execute any example directly, for example:

python examples/01_mne_integration.py

or

python examples/02_pure_jax_fitting.py

---

Advanced Examples

The repository also includes additional examples that are not part of the online gallery:

• Validation experiments comparing amica with the original MATLAB implementation.

• HPC / Slurm templates for running AMICA on computing clusters.

• Jupyter notebooks demonstrating complete EEG workflows.

These can be found in the examples/validation/, examples/cluster/, and notebook files in the repository.