bobthecat

library(rbenchmark)
res <- benchmark( cosine(mat),
                  cosineRcpp(mat),
                  columns=c("test", "replications", "elapsed",
                           "relative", "user.self", "sys.self"),
                  order="relative",
                  replications=1000)

res
             test replications elapsed relative user.self sys.self

bobthecat

library(twitteR)
library(ggplot2)
library(grid)

# download all that you can
pol <- userTimeline('BeijingAir', n=3200)
length(pol)
# 3200

myGrep <- function(x){

bobthecat

library(twitteR)
library(ggplot2)
library(grid)

# download all that you can
pol <- userTimeline('BeijingAir', n=3200)
length(pol)
# 3200

myGrep <- function(x){

bobthecat

library(GEOquery); library(RankProd); library(mouse4302.db)
## Download the data from GEO
gse12499 <- getGEO('GSE12499',GSEMatrix=TRUE)
e <- exprs(gse12499[[1]])
dim(e)
[1] 45101    10

bobthecat

### R code from vignette source 'Presentation.Rnw'
### Encoding: UTF-8

###################################################
### code chunk number 1: init
###################################################
options(width=60)


###################################################

bobthecat

### R code from vignette source 'Presentation_2.Rnw'
### Encoding: UTF-8

###################################################
### code chunk number 1: init
###################################################
options(width=60)


###################################################

bobthecat

library(xlsx)
library(googleVis)
# I downloaded the Excel file, cleaned the headers and worked a bit 
# the column title.
da <- read.xlsx("~/Downloads/religion.xlsx", sheetName=1)
rownames(da) <- da$COUNTRY.
da <- da[,-1]
religion <- data.frame(country=rep(rownames(da), 3), 
  year=c(rep(2007, dim(da)[1]), rep(2009, dim(da)[1]), rep(2010, dim(da)[1])),
  GRI=c(da$GRI_2007, da$GRI_2009, da$GRI_2010), 

bobthecat

library(GEOquery)
## Download the data from GEO
GDS3716 <- getGEO('GDS3716')
# transform the GDS to and expressionSet
eset <- GDS2eSet(GDS3716,do.log2=TRUE)
phenoData <- pData(eset)
# keep only the ER+ and ER-
samples <- phenoData$sample[grep("ER", phenoData$specimen)]
# subsetting the expressionSet
eset <- eset[,samples]

bobthecat

## EXTRACTING CLASS LABELS
classLabel <- sub("^ER(.*) breast cancer", "\\1", grep("ER", phenoData$specimen, value=T))
classLabel
[1] "-" "-" "-" "-" "-" "-" "-" "-" "-" "+" "+" "+" "+" "+" "+" "+" "+" "+"

## COMPUTING P-VALUE DISTRIBUTION
minus = which(classLabel=="-")
plus = which(classLabel=="+")
p <- apply(e, 1, function(x){t.test(as.numeric(x[minus]), as.numeric(x[plus]))$p.value})

bobthecat

## LOADING LIBRARIES FOR PARALLEL PROCESSING
library(doMC)
ncore = multicore:::detectCores()
registerDoMC(cores = ncore)

## COMPUTING THE RANDOM P-VALUE DISTRIBUTION
# How many random sampling
R=100
# Shuffling the sample labels and recomputing the p-value each time
p.rand <- foreach(i = 1:dim(e)[1], .combine=rbind) %dopar% {