mean_correlation_for_mask_set: Given a precomputed matrix of masked correlations, gather...
In mkuhn/maskedCorr:

Given a precomputed matrix of masked correlations, gather correlations for the specified mask set.

1	mean_correlation_for_mask_set(masks, correlations, Nbits, masks_filter = NULL)

`masks`	vector of masks (length N) from which to take correlations
`correlations`	NxN matrix of correlations, where each matrix element is a vector with entries for each masked correlation
`Nbits`	optional: number of bits, must be supplied if masks_filter is given
`masks_filter`	optional: set of masks (of the comparison, not the input mask) for which to return correlations, supply as integer-encoded mask (not as vector)

A list of two elements: the average correlation, and a sorted vector of the encountered combination of masks. It is possible that the given masks and correlations are partially not compatible, Then, the average will be computed based on those mask combinations that contain a correlation. (E.g. in the mask triple "A, AB, B" the average will be computed from two correlations (A/A, B/B), as it is not possible to calculate a correlation between A and B.)

mkuhn/maskedCorr documentation built on May 23, 2019, 2:03 a.m.