A python implementation of the mosaic permutation testing framework.
To install mosaicperm, just use pip:
pip install mosaicperm
Documentation and tutorials are available at https://mosaicperm.readthedocs.io/.
Below, we give a simple example showing how to use mosaicperm to test whether a set of factor exposures explain the correlations among a matrix of outcomes variables.
import numpy as np
import mosaicperm as mp
# synthetic outcomes and exposures
n_obs, n_subjects, n_factors = 100, 200, 20
outcomes = np.random.randn(n_obs, n_subjects)
exposures = np.random.randn(n_obs, n_subjects, n_factors)
# example of missing data
outcomes[0:10][:, 0:5] = np.nan
exposures[0:10][:, 0:5] = np.nan
# fit mosaic permutation test
mpt = mp.factor.MosaicFactorTest(
outcomes=outcomes,
exposures=exposures,
test_stat=mp.statistics.mean_maxcorr_stat,
)
print(mpt.fit().summary())
# produce a time series plot of this analysis
mpt.fit_tseries(
nrand=100, n_timepoints=20,
).plot_tseries()
See the documentation for more details.
If you use mosaicperm in an academic publication, please consider citing our paper:
@article{mosaic2024,
author = {Spector, Asher and Barber, Rina Foygel and Hastie, Trevor and Kahn, Ronald N. and Candès, Emmanuel},
title = {The mosaic permutation test: an exact and nonparametric goodness-of-fit test for factor models},
date = {2024},
annotation = {2024f},
eprint = {2404.15017},
eprintclass = {stat.ME},
eprinttype = {arXiv},
url={https://arxiv.org/abs/2404.15017},
}
