Generate a matrix of average counts given replicates

[vidalrmartinez]

Jason would like a feature which takes the sample columns of a given input matrix and averages these columns based on the given replicates (sample names provided by the '--input' option). Each set of replicates per line determine the sample columns to join. The count for the joined columns is the average of the columns.

For example,

a given matrix with 50 sample columns will result in a 25 column matrix if 2 replicates are used. If a pair of replicates is specified BB.3_1X_NS30 and BB.18_1X_NS30, then for every probe the count in the new matrix in column BB.3_1X_NS30~BB.18_1X_NS30 is the average of the two counts from the same probe in the old matrix in column BB.3_1X_NS30 and BB.18_1X_NS30.

[jtladner]

This looks pretty good Vidal, but I want the user to specify the name of the new column in the output file. So, as an example, here is a hypothetical input file. Four sample columns would be included in the output, with names specified by the first column in this file. In one case, the input sample column will be included as is in the output because only a single replicate is provided. In two cases, the new columns will be generated by averaging two replicates each. And in one case, three replicates would be averaged.

exampleInputFile.txt
.

[jtladner]

@jdr479 - Here are example input and output files you can use for testing, along with the example names file I already posted ("exampleInputFile.txt").
output_score_matrix.txt
input_score_matrix.txt

[Num0Programmer]

Closed because implementation was pushed to master branch.