Matt:
> from collections import defaultdict
>
> def get_hist(file_name):
> hist = defaultdict(int)
> f = open(filename,"r")
> for line in f:
> vals = line.split()
> val = int(vals[0])
> try: # don't look to see if you will cause an error,
> # just cause it and then deal with it
> cnt = int(vals[1])
> except IndexError:
> cnt = 1
> hist[val] += cnt
> return hist
But usually in tight loops exceptions slow down the Python code, so
this is quite faster (2.5 times faster with Psyco, about 2 times
without, with about 30% of lines with a space in it):
import psyco
from collections import defaultdict
def get_hist(file_name):
hist = defaultdict(int)
for line in open(file_name):
if " " in line:
pair = line.split()
hist[int(pair[0])] += int(pair[1])
else:
hist[int(line)] += 1
return hist
psyco.bind(get_hist)
It doesn't work if lines may contain spurious spaces...
Bye,
bearophile
--
http://mail.python.org/mailman/listinfo/python-list
