X-Git-Url: https://git.donarmstrong.com/?a=blobdiff_plain;ds=sidebyside;f=EBSeq%2Frsem-for-ebseq-generate-ngvector-from-clustering-info;fp=EBSeq%2Frsem-for-ebseq-generate-ngvector-from-clustering-info;h=312dbfa33c77339b530e67362e14e391e75f4259;hb=9c2e46183a19d661f0a618a8eabe8ce1f6a8e2d6;hp=0000000000000000000000000000000000000000;hpb=7777bfc71742d4e239d1297d2e3de058895f4ce1;p=rsem.git

diff --git a/EBSeq/rsem-for-ebseq-generate-ngvector-from-clustering-info b/EBSeq/rsem-for-ebseq-generate-ngvector-from-clustering-info
new file mode 100755
index 0000000..312dbfa
--- /dev/null
+++ b/EBSeq/rsem-for-ebseq-generate-ngvector-from-clustering-info
@@ -0,0 +1,18 @@
+#!/usr/bin/env Rscript
+
+argv <- commandArgs(TRUE)
+if (length(argv) != 2) {
+  cat("Usage: rsem-for-ebseq-generate-ngvector-from-clustering-info input_file output_file\n")
+  q(status = 1)
+}
+
+data <- read.table(file = argv[1], stringsAsFactors = F)
+idx <- data[,2] >= 0
+kmr <- kmeans(data[idx, 2], 3)
+order <- order(kmr$centers)
+
+ngvec <- rep(0, length(idx))
+ngvec[idx] <- order[kmr$cluster]
+ngvec[!idx] <- 3
+
+write.table(ngvec, file = argv[2], row.names = F, col.names = F)