causality: Accessor function for the 'causality' information in a...

In bhklab/mRMRe: Parallelized Minimum Redundancy, Maximum Relevance (mRMR)

Description

The causality data is compute using the co-information lattice algorithm on each V-structure (feature, target, feature). Given that this procedure is computed for each pair of features, the minimum result is kept. A negative score indicates putative causality of the feature to the target.

Usage

## S4 method for signature 'mRMRe.Filter'
causality(object)
## S4 method for signature 'mRMRe.Network'
causality(object)

Arguments

object

a mRMRe.Filter or mRMRe.Network object.

Author(s)

Nicolas De Jay, Simon Papillon-Cavanagh, Benjamin Haibe-Kains

Examples

set.thread.count(2)
data(cgps)
feature_data <- mRMR.data(data =  data.frame(cgps.ge))
filter <- mRMR.classic("mRMRe.Filter", data = feature_data, target_indices = 3:5,
						feature_count = 2)
causality(filter)

bhklab/mRMRe documentation built on Sept. 3, 2021, 10:50 p.m.