Mercurial > repos > melissacline > ucsc_cancer_utilities
diff mergeGenomicMatrixFiles.py @ 14:d0b8c8eee9d5
Committing changes that hadn't made it in yet, merge hell
author | melissacline |
---|---|
date | Tue, 10 Mar 2015 19:32:14 -0700 |
parents | 30aab34424a9 |
children | 1d83dbbee373 |
line wrap: on
line diff
--- a/mergeGenomicMatrixFiles.py Tue Mar 10 19:29:10 2015 -0700 +++ b/mergeGenomicMatrixFiles.py Tue Mar 10 19:32:14 2015 -0700 @@ -1,21 +1,27 @@ #!/usr/bin/env python +import argparse import string,os,sys -def header (samples, infile): - fin= open(infile,'r') +def header (samples, sourceFiles, infile, labelThisFile): + if labelThisFile == None: + labelToUse = infile + else: + labelToUse = labelThisFile + fin= open(infile, 'U') #header, samples newSamples = string.split(string.strip(fin.readline()),'\t')[1:] for sample in newSamples: if sample not in samples: samples[sample]= len(samples) + sourceFiles[sample] = labelToUse fin.close() return def process(genes, samples, dataMatrix, infile): maxLength= len(samples) - fin= open(infile,'r') + fin= open(infile,'U') #header newSamples = string.split(string.strip(fin.readline()),'\t') @@ -41,7 +47,17 @@ fin.close() return -def outputMatrix(dataMatrix, samples, genes, outfile): + +def outputSourceMatrix(sourceData, outputFileName): + fout = open(outputFileName, "w") + fout.write("Sample\tSource\n") + for thisSample in sourceData.keys(): + fout.write("%s\t%s\n" % (thisSample, sourceData[thisSample])) + fout.close() + return + + +def outputMergedMatrix(dataMatrix, samples, genes, outfile): fout = open(outfile,"w") maxLength= len(samples) sList=[] @@ -66,23 +82,38 @@ return if __name__ == '__main__' : - if len(sys.argv[:]) <4: - print "python mergeFilesByColumn.py output inputfile(s)" - print "**********memory intensive, not for very genomic data with hugo number of probes" - print "this is merging data A+B=C\n" - sys.exit() - - inFiles = sys.argv[2:] - outfile = sys.argv[1] + # + # The input files to this script are two or more matrices, in which + # columns represent samples and rows represent genes or measurements. + # There are two output files: outMergedData contains the input data merged + # into a single matrix, and outSourceMatrix is a two-column matrix + # indicating which file each sample (or column label) came from. This + # assumes that each sample came from at most one file. + # + parser = argparse.ArgumentParser() + parser.add_argument("inFileA", type=str, help="First input file") + parser.add_argument("inFileB", type=str, help="Second input file") + parser.add_argument("outMergedData", type=str, + help="Filename for the merged dataset") + parser.add_argument("outSourceMatrix", type=str, + help="""Filename for a Nx2 matrix that indicates + the source file of each column""") + parser.add_argument("--aLabel", type=str, default=None, + help="User-friendly label for the first input file") + parser.add_argument("--bLabel", type=str, default=None, + help="User-friendly label for the second input file") + args = parser.parse_args() genes={} samples={} + sourceFiles = {} dataMatrix=[] - for infile in inFiles: - header (samples, infile) + header(samples, sourceFiles, args.inFileA, args.aLabel) + header(samples, sourceFiles, args.inFileB, args.bLabel) - for infile in inFiles: - process(genes, samples, dataMatrix, infile) + process(genes, samples, dataMatrix, args.inFileA) + process(genes, samples, dataMatrix, args.inFileB) - outputMatrix(dataMatrix, samples, genes, outfile) + outputSourceMatrix(sourceFiles, args.outSourceMatrix) + outputMergedMatrix(dataMatrix, samples, genes, args.outMergedData)