diff options
| author | Patrick Simianer <simianer@cl.uni-heidelberg.de> | 2012-11-05 15:29:46 +0100 |
|---|---|---|
| committer | Patrick Simianer <simianer@cl.uni-heidelberg.de> | 2012-11-05 15:29:46 +0100 |
| commit | 6f29f345dc06c1a1033475eac1d1340781d1d603 (patch) | |
| tree | 6fa4cdd7aefd7d54c9585c2c6274db61bb8b159a /gi/evaluation/entropy.py | |
| parent | b510da2e562c695c90d565eb295c749569c59be8 (diff) | |
| parent | c615c37501fa8576584a510a9d2bfe2fdd5bace7 (diff) | |
merge upstream/master
Diffstat (limited to 'gi/evaluation/entropy.py')
| -rw-r--r-- | gi/evaluation/entropy.py | 38 |
1 files changed, 0 insertions, 38 deletions
diff --git a/gi/evaluation/entropy.py b/gi/evaluation/entropy.py deleted file mode 100644 index ec1ef502..00000000 --- a/gi/evaluation/entropy.py +++ /dev/null @@ -1,38 +0,0 @@ -#!/usr/bin/env python - -import sys, math, itertools, getopt - -def usage(): - print >>sys.stderr, 'Usage:', sys.argv[0], '[-s slash_threshold] input file' - sys.exit(0) - -optlist, args = getopt.getopt(sys.argv[1:], 'hs:') -slash_threshold = None -for opt, arg in optlist: - if opt == '-s': - slash_threshold = int(arg) - else: - usage() -if len(args) != 1: - usage() - -infile = open(args[0]) -N = 0 -frequencies = {} - -for line in infile: - - for part in line.split('||| ')[1].split(): - tag = part.split(':',1)[1] - - if slash_threshold == None or tag.count('/') + tag.count('\\') <= slash_threshold: - frequencies.setdefault(tag, 0) - frequencies[tag] += 1 - N += 1 - -h = 0 -for tag, c in frequencies.items(): - h -= c * (math.log(c, 2) - math.log(N, 2)) -h /= N - -print 'entropy', h |
