makePhaseOneOrderedRM: Order rules according to phase one importance ranking

Description Usage Arguments Value Examples

Description

Order rules according to phase one importance ranking

Usage

1
makePhaseOneOrderedRM(D, rm.start, spr, Q, trn, n.splits, shouldPrint)

Arguments

D

Binary matrix of N events and M samples

rm.start

Starting binary rule matrix (i.e., rule library)

spr

Random rule sets per rule in each phase one iteration. Default is 40.

Q

Penalty matrix, negative log of passenger probability matrix.

trn

Target rule number for stopping iterating. Default is 16.

n.splits

number of splits for parallelization. Default is all available cpus.

shouldPrint

Print progress updates? Default is TRUE

Value

binary rule matrix ordered by phase one importance ranking

Examples

1
2
3
4
5
6
data(skcm)
list2env(skcm.list,envir=globalenv())
Q <- log10(P)
rm.full <- buildRuleLibrary(D,rule.thresh = 0.06) # Rule library matrix, dimension: 36s x 71
rm.ordered <- makePhaseOneOrderedRM(D,rm.full,spr = 1,Q,trn = 34,shouldPrint = TRUE)
# note, for real applications, spr should be at least 40.

crso documentation built on July 7, 2019, 5:02 p.m.