##################################
#                                #
# Last modified 2019/08/16       # 
#                                #
# Georgi Marinov                 #
#                                # 
##################################

import sys
import string

def run():

    if len(sys.argv) < 4:
        print 'usage: python %s input labelfields valuefields outputfilename' % sys.argv[0]
        print '       valuefields format: either comma separated, or start:end (including start and end, 0-based)'
        sys.exit(1)
    
    input = sys.argv[1]
    outfilename = sys.argv[4]

    fields = sys.argv[2].split(',')
    labelFields=[]
    for f in fields:
        labelFields.append(int(f))
    labelFields.sort()

    print labelFields

    valueFields=[]
    if ':' in sys.argv[3]:
        fields = sys.argv[3].split(':')
        start = int(fields[0])
        end = int(fields[1])
        for f in range(start,end+1):
            valueFields.append(f)
    else:
        fields = sys.argv[3].split(',')
        for f in fields:
            valueFields.append(int(f))
    valueFields.sort()

    print valueFields

    DataDict={}
    for ID in valueFields:
        DataDict[ID] = 0

    outfile = open(outfilename, 'w')

    L=0.0
    linelist = open(input)
    for line in linelist:
        fields=line.replace('\x00','').strip().split('\t')
        if line.startswith('#') or line.startswith('tracking_id'):
            if len(fields) < max(valueFields):
                continue
            outline = ''
            for ID in labelFields:
                outline = outline + fields[ID] + '\t'
            for ID in valueFields:
                outline = outline + fields[ID] + '\t'
            outfile.write(outline + '\n')
            continue
        L+=1
        for ID in valueFields:
            if fields[ID] != 'nan':
                DataDict[ID] += float(fields[ID])

    outline = ''
    for ID in labelFields:
        outline = outline + '\t'
    outline = outline[:-1]

    for ID in valueFields:
        outline = outline + '\t' + str(DataDict[ID]/L)

    outfile.write(outline + '\n')

    outfile.close()
   
run()