combineDists

#!/usr/bin/python

import sys
import numpy as np
from scipy import stats

def getOptions():
    import argparse

    # create the top-level parser
    description = ("Takes a bunch of distance matrices and outputs the average")
    parser = argparse.ArgumentParser(description = description)
    parser.add_argument('DIST_FILE', action='store', nargs='+',
                            help='Distant matrices')
    parser.add_argument('-E', '--error', action="store_true",
            default=False,
            dest='error',
            help='Compute the standard error of the values')
    return parser.parse_args()

options = getOptions()

files = options.DIST_FILE

def readDist(infile):
    b = True
    
    d = {}
    for l in open(infile):
        s = l.rstrip().split('\t')
        
        if b:
            orgs = s[1:]
            b = False
            continue
        
        o = s[0]
        values = s[1:]
        d[o] = d.get(o, {})
        for x in orgs:
            d[o][x] = float( values[orgs.index(x)] )
            
    return d

d = {}
for f in files:
    t = readDist(f)
    
    for o in t:
        d[o] = d.get(o, {})
        for o1 in t[o]:
            d[o][o1] = d[o].get(o1, [])
            d[o][o1].append(t[o][o1])
            
orgs = set(d.keys())

print '\t'.join( [''] + sorted(orgs) )
for o in sorted(orgs):
    if options.error:
        print('\t'.join( [o] + [str(round(stats.sem(np.array(d[o][o1])),5)) if o1 in d[o] else '0' for o1 in sorted(orgs)] ))
    else:
        print('\t'.join( [o] + [str(round(np.array(d[o][o1]).mean(),5)) if o1 in d[o] else '0' for o1 in sorted(orgs)] ))