principal_curve_analysis_tsne_d3.R

#
# R file to do principal curve analysis on tsne output
#

library(princurve)


args <- commandArgs()

#ARGS
baseName <- args[6]


infile <- paste(getwd(), "sample_data", baseName, "intermediate_files",
                paste(baseName, "_tsne_d3.csv",sep=""), sep="/")
pcvout <- paste(getwd(), "sample_data", baseName, "intermediate_files",
                paste(baseName, "_tsne_d3_pcv.csv",sep=""), sep="/")
lambdaout <- paste(getwd(), "sample_data", baseName, "intermediate_files",
                   paste(baseName, "_tsne_d3_lambda.csv",sep=""), sep="/")

selectedCellInfile <- paste(getwd(), "sample_data", baseName, "intermediate_files",
                            paste(baseName, "_initial_cell.txt",sep=""), sep="/")




x <- read.csv(file=infile, header=TRUE, sep=",")


y <- data.matrix(x)


# code for not initializing the starting point lowess version
fitpc <- principal.curve(y, plot = TRUE, smoother = "lowess", maxit = 200)

write.table(file=pcvout, sep=",", fitpc$s)
write.table(file=lambdaout, sep=",", fitpc$lambda)