script 1: convert arff to vw

try to apply vowpal wabbit on malicious websites data. this is a useful script can convert a

arff(weka) format to vw required format. a perl script posts online actually not working at all.

#!/usr/bin/python
import sys

filename = sys.argv[1]

f = open(filename, "r")
fout = open("total1.vw", 'w')
for line in f:
  vwline = ""
  if len(line)<4 or line.startswith("@"):
    continue
  strs = line.split(',')
  strs = map(lambda x: x.strip(), strs)
  vwline = strs[-1] + " |f "
  for i,e in enumerate(strs[:-1]):
    vwline += str(i) + ':' + e + " "
  
  fout.write(vwline)
  fout.write('\n')
 
f.close()
fout.close()