gistfile1.py

#! /usr/bin/env python
from pandas import *
from numpy import *
import sys
from itertools import * 
from math import exp 

def logistic(num):
    return (exp(num)/(1+exp(num)))

fromvw = sys.argv[1]  # blah.vw , the file that has the example vowpal learns from
preddat = sys.argv[2] # blah-run.dat , prediction output file from vw blah-run
target = sys.argv[3] # pairedLabelVs.csv , csv file with prediction and 



fl = open(target,"w")
fl.write("isclick, prediction\n")
# i have written this code but I have not tested it
labchunks = read_csv(fromvw, sep=" ",header=None,chunksize = 100000) # 
predchunks= read_table( preddat ,header=None, names=["prediction"],chunksize = 100000 )
print labchunks
print predchunks
for labels,preds in izip(labchunks,preddat):
#     print labels
#     print predd
    # currentrange = min( len(labels.ix[:,0]), len(predd.ix[:,0]))
    # print preds 
    for lbl,prd  in zip( labels.ix[:,"X.1"] ,preds.prediction):
            fl.write(str(lbl) + "," +str(logistic(prd))+ "\n" )
    fl.flush()
fl.close()