test.stat: Gene test statistics dataset
In BANFF: Bayesian Network Feature Finder

The simulated test statistics on each gene node. With 10% randomly selected as missing.

1	data(test.stat)

A vector of length=#genes, the value is the observed test statistics for each gene.

We first simulate the underlying true feature labels by 25% down-regulate, 50% null and 25% up-regulate. The pre-defined distribution for each regulation group is normal: the down-regulate genes are distributed as a normal with mean=-0.5 and sd=0.2, the null genes are distributed as a normal with mean=0 and sd=0.2, the up-regulated genes are distributed as a normal with mean=0.5 and sd=0.2. For the second step, we randomly knock out 10% genes as missing nodes.

BANFF documentation built on May 29, 2017, 11:59 a.m.