v0.4.1

Full Changelog: v1.0.0...v1.0.1

v0.4.0

Monte-Carlo permutation-testing cluster analysis (maxT algorithm) or FDR-corrected analysis working for comparisons (i.e., comparing two groups or two levels of a repeated-measure factor) and associations (i.e., correlations between a continuous variable and each physiological variable).
Fully working for 2d data (e.g., time and channels like ERP). This release should be able to handle most wanted ERP-like analyses.
Not fully tested for 3d data

v0.3.0

This release contains a new plotting function that flattens 3D scalp sensor locations. Currently, this function is used by default when evaluating channel neighborhoods. Otherwise it can be used on its own to make simple scalp maps.

v0.2.0

Monte-Carlo permutation-testing cluster analysis (maxT algorithm) or FDR-corrected analysis working for comparisons (i.e., comparing two groups or two levels of a repeated-measure factor) and associations (i.e., correlations between a continuous variable and each physiological variable)
Note: Tested on a couple of EEG datasets, but not thoroughly checked yet. Next release coming soon with more features.

v0.1.0

This version allows Monte-Carlo based permutation or H0-bootstrap cluster-level statistics. Clusters are defined through proximity in a 3d space (two continuous dimensions, such as time and frequency and one discrete channel dimension) or a subspace. The statistics are for comparing groups, comparing conditions, or correlations.