multisim-neuro 0.1

A package for simulating time resolved data with multivariate effects

MultiSim

Simulate realistic MEG/EEG data with ground-truth multivariate effects

Installation

pip install multisim-neuro

Or clone and install the development version:

git clone https://github.com/AlexLepauvre/multisim-neuro
cd multisim-neuro
pip install -e .[full]

Features

• Design-driven simulation: specify arbitrary experimental design matrices (conditions, interactions, parametric regressors).

• Time‑locked effects: inject multivariate patterns in chosen time windows with exact control over effect size (Mahalanobis distance).

• Noise modeling: control within‑subject sensor noise and between‑subject variability.

• Spatial covariance: simulate correlated sensors or modes via any user‑supplied covariance matrix.

• MNE & EEGLAB export: seamless conversion to mne.EpochsArray or EEGLAB .set files.

• Power analysis: predict group‑level statistical power from trials, subjects, and effect sizes.

Quickstart

import numpy as np
from multisim import Simulator

# 1. Create a simple design: two conditions, 100 trials
X = np.vstack([np.zeros(100), np.ones(100)]).T  
t_win   = np.array([[0.1, 0.3]])    # effect between 100–300 ms
effects = np.array([1])             # effect on condition 1

# 2. Instantiate simulator
sim = Simulator(
    X, noise_std=1.0, n_channels=64, n_subjects=10,
    tmin=-0.2, tmax=0.8, sfreq=250,
    t_win=t_win, effects=effects,
    effect_size=[0.5],               # multivariate d′ = 0.5
    intersub_noise_std=0.1           # between-subject σ
)

# 3. Export to MNE Epochs and decode
epochs_list = sim.export_to_mne()
# ... run your decoding pipeline on epochs_list ...

API

After installation, see full class and method documentation at: https://alexlepauvre.github.io/multisim-neuro/index.html

You can find extensive tutorials at: https://alexlepauvre.github.io/multisim-neuro/tutorial/index.html

Customization

• Spatial pattern: pass a custom covariance matrix or weight vector to concentrate signal in subsets of channels.

• Temporal kernel: supply any causal kernel to shape the time-course of effects.

• Effect size: Simulate multivariate effect at particular effect size to test the ability of your pipelines to retrieve it given the number of subjects and trials for each subject

How to cite us:

If you use the scripts found in this repository, you can use the DOI provided by Zenodo to cite us. And here is a bibtex:

@article{LepauvreEtAl2024,
  title = {MultiSim},
  author = {Lepauvre, Alex and Chu, Qian and Zeidman, Peter and Melloni, Lucia},
  year = {2025},
  doi = {TBA},
}