FamilyRank-package {FamilyRank}

Algorithm for Ranking Predictors Using Graphical Domain Knowledge

Description

Grows families of features by selecting features that maximize a weighted score calculated from empirical feature scores and graphical knowledge. The final weighted score for a feature is determined by summing a feature's family-weighted scores across all families in which the feature appears.

Details

The DESCRIPTION file:

Package:	FamilyRank
Type:	Package
Title:	Algorithm for Ranking Predictors Using Graphical Domain Knowledge
Version:	1.0
Date:	2021-01-24
Author:	Michelle Saul
Maintainer:	Michelle Saul <[e-mail]>
Description:	Grows families of features by selecting features that maximize a weighted score calculated from empirical feature scores and graphical knowledge. The final weighted score for a feature is determined by summing a feature's family-weighted scores across all families in which the feature appears.
License:	GPL
Imports:	Rcpp (>= 1.0.6), plyr (>= 1.8.6), stats (>= 3.6.0)
LinkingTo:	Rcpp, RcppArmadillo
LazyData:	true

Index of help topics:

FamilyRank-package      Algorithm for Ranking Predictors Using
                        Graphical Domain Knowledge
createCase              Simulate Cases
createControl           Simulate Control
createData              Simulate Data
createGraph             Simulate Graph
familyRank              Feature Ranking with Family Rank
grow                    Grow Families
indexFeats              Re-index features
rbinorm                 Bimodal Normal Distribution

The main function is familyRank.

Author(s)

Michelle Saul

Maintainer: Michelle Saul <[e-mail]>

References

ADD REFERENCE