binom_2proptest: P value detection for TSPs

In KellyCahill/rfTSP: robust classification and feature selection for binary/multi outcomes

Description

P value selection criteria for optimal number of k TSPs

Usage

 
binom_2proptest(j, train, N, controls, case, ind) 

Arguments

ind	a matrix consisting of 2 columns and p rows of gene pairs for all gene pairs, not just those returned by getKtsp.
j	the gene pair index for the pair for a p-value to be calculated. Will be looped over all pairs, see example.
case	index of case subjects corresponding to columns in "train".
control	control index of control subjects corresponding to columns in "train".
N	N total number of training subjects, should equal number of columns in "train".

Value

pvalue from a rigorous test of TSP pattern intensity for each gene pair.

Author(s)

Kelly Cahill, Masaki Lin, Silvia Liu

KellyCahill/rfTSP documentation built on May 21, 2019, 1:41 p.m.