# efg, Stowers Institute, Jan 2006

switch(.Platform$OS.type,
         windows = OS.Prefix <- "U:",
         unix    = OS.Prefix <- "/n/projects",
         stop("unsupported OS platform")
      )

basepath <- paste(OS.Prefix,
                  "/camda/2006/gene_expression_values", sep="")


Inclusion <- read.csv("U:/camda/2006/clinical_data/NoExclusions.csv", as.is=TRUE)
txt.list <- list.files(path=basepath, pattern=".txt")
txt.CLUSTER <- rep("EXCLUDED", length(txt.list))

for (k in 1:length(txt.list))
{
  filename <- txt.list[k]
  index <- which( substr(filename,1,8) == Inclusion$ABTID )
  if (length(index) > 0) txt.CLUSTER[k] <- Inclusion$CLUSTER[index]
}

write.csv(data.frame(ABTID=txt.list, CLUSTER=txt.CLUSTER),
          file="ArrayAssignCluster.csv", row.names=FALSE)

table(txt.CLUSTER)


write.list <- function(target)
{
  ABTID <- data.frame(ABTID=txt.list[ txt.CLUSTER == target] )
  write.csv(ABTID, row.names=FALSE, file=paste(target, ".csv", sep=""))
}

write.list("Least")
write.list("Middle")
write.list("Worst")
write.list("EXCLUDED")