Difference between revisions of "GGBsnippet.R"
From Organic Design wiki
m |
m (# {{R}}) |
||
(One intermediate revision by the same user not shown) | |||
Line 1: | Line 1: | ||
+ | # {{R}} | ||
# p (genes) by n (slides) matrix | # p (genes) by n (slides) matrix | ||
# m=p keeping fdr notation | # m=p keeping fdr notation | ||
Line 27: | Line 28: | ||
# Graphical check | # Graphical check | ||
+ | hist(kDE, prob=TRUE) | ||
+ | lines(density(kDE)) | ||
+ | |||
+ | hist(kEE, prob=TRUE) | ||
+ | lines(density(kEE)) | ||
+ | |||
+ | Xbar <- rowMeans(log2(X)) | ||
+ | hist(Xbar, breaks=30, prob=TRUE) | ||
+ | lines(density(Xbar)) | ||
pairs(log2(X)) | pairs(log2(X)) |
Latest revision as of 00:38, 6 June 2007
Code snipits and programs written in R, S or S-PLUS
- p (genes) by n (slides) matrix
- m=p keeping fdr notation
m <- p <- 1000 nreps <- 5
n <- nreps * 2 pi0 <- 0.95
m1 <- round(m * (1-pi0)) m0 <- p - m1
params <- c(2.74886, 1.36546, 4.12844) # "IPTG-a parameters"
a.shape <- params[1] a0.shape <- params[2] scale <- params[3]
DEscales <- rgamma(2*m1, shape=a0.shape, rate=scale) EEscales <- rgamma(m0 , shape=a0.shape, rate=scale)
scales <- c(rep(DEscales, each=nreps), rep(EEscales, each=2*nreps)) X <- rgamma(n* p, a.shape, rate=scales) dim(X) <- c(n,p)
X <- t(X)
- Graphical check
hist(kDE, prob=TRUE) lines(density(kDE))
hist(kEE, prob=TRUE) lines(density(kEE))
Xbar <- rowMeans(log2(X)) hist(Xbar, breaks=30, prob=TRUE) lines(density(Xbar)) pairs(log2(X))