Calculates the point wise mutual information — rs_pointwise_mutual

Calculates the pointwise mutual information (can be also normalised) across all columns of the data. This can be used to identify (dis)similar samples based on Boolean characteristics.

Usage

rs_pointwise_mutual_info(x, normalise)

Arguments

x: Logical matrix. The columns represent features and the rows represent samples
normalise: Shall the normalised pointwise mutual information be returned.

Value

The (normalised) pointwise mutual information matrix between the columns. Pairs without co-occurrence are -Inf.