API Reference

Factor models (mosaicperm.factor)

MosaicFactorTest(outcomes, exposures, test_stat)	Mosaic test for factor models with known exposures.
MosaicBCV(new_exposures, *args, **kwargs)	Mosaic factor test based on mosaicperm.statistics.adaptive_mosaic_bcv_stat()
ols_residuals(outcomes, exposures)	Computes residuals via cross-sectional OLS.

Panel data (mosaicperm.panel)

MosaicPanelInference(*args, **kwargs)	Mosaic permutation-based inference for linear models in panel data.
MosaicPanelTest(outcomes, test_stat, ...[, ...])	Initialize a MosaicPanelTest for panel data analysis.
QuadraticMosaicPanelTest(*args[, method, ...])	Mosaic permutation test for panel data using quadratic test statistics.
ols_residuals_panel(outcomes, covariates)	Computes residuals via OLS (NOT cross-sectional OLS).

Combining results (mosaicperm.core)

combine_mosaic_tests(mosaic_objects[, ...])	Combines results across MosaicPermutationTest objects.
combine_mosaic_tests_tseries(mosaic_objects)	Combines time-series results across MosaicPermutationTest objects.

Test statistics (mosaicperm.statistics)

mean_abscorr_stat(residuals, **kwargs)	Mean absolute correlation statistic.
mean_maxcorr_stat(residuals, **kwargs)	Mean maximum (absolute) correlation statistic.
quantile_maxcorr_stat(residuals[, qs])	Quantiles of maximum (absolute) correlation statistic.
mosaic_bcv_stat(residuals, new_exposure, tiles)	Out-of-sample R^2 measuring improvement for an augmented model containing an additional exposure.
adaptive_mosaic_bcv_stat(residuals, ...[, mus])	Computes a mosaic BCV statistic for several augmented models.
approximate_sparse_pcas(Sigma[, quantiles])	Performs approximate sparse pca.
active_subset(residuals[, subset, min_std])	Returns subjects whose residual stdev is above min_std.

Tilings (mosaicperm.tilings)

Tiling(tiles[, check_valid])	A class to store tilings used in mosaic tests.
default_factor_tiles(exposures[, n_obs, ...])	Computes default tiling for factor models.
random_tiles(n_obs, n_subjects, nbatches, ...)	Partitions outcomes into nbatches x ngroups random tiles.
check_valid_tiling(tiles)	Checks validity of a tiling.

Other core functions (mosaicperm.core)

compute_adaptive_pval(statistic, null_statistics)	Computes an adaptive p-value based on one or more test statistics.
MosaicPermutationTest()	Generic class meant for subclassing.