R/findOutliers.R

In aalmasaud/TDW:

findOutliers <- function(mydatafile,dataColumn,...) {
dataFile <- data.frame(mydatafile)
#iqr1 <- aggregate(dataColumn ~ Hour + Lane,data=dataFile,FUN=function(x) quantile(x, probs = 0.25)-IQR(x)*2)
iqr1 <- aggregate(dataFile[,dataColumn],by=list(Hour=dataFile$Hour,Lane=dataFile$Lane),FUN=function(x) quantile(x, probs = 0.25)-IQR(x)*2)
#iqr3 <- aggregate(dataColumn ~ Hour + Lane,data=dataFile,FUN=function(x) quantile(x, probs = 0.75)+IQR(x)*2)
iqr3 <- aggregate(dataFile[,dataColumn],by=list(Hour=dataFile$Hour,Lane=dataFile$Lane),FUN=function(x) quantile(x, probs = 0.75)+IQR(x)*2)
stats<-merge(iqr1,iqr3,by=c("Hour","Lane"))

new <- merge(dataFile,stats,by=c("Hour","Lane"))

names(new)<- gsub("\\.","_",names(new))

rOutliers <- filter2(new,paste(dataColumn," < x_x OR ",dataColumn," > x_y",sep=""))
rWithoutOutliers1 <- filter2(new,paste(dataColumn," > x_x AND ",dataColumn," < x_y",sep=""))
rWithoutOutliers <- removeColumn(rWithoutOutliers1,'["x_x","x_y"]')
return(rWithoutOutliers)
}