comparison g_collocation.py @ 1:fb617586f4b2 draft

planemo upload commit a81826fe44f09a3710a35c183aa88b745aeec064-dirty

0:e991d4e60c17

    args = the_parser.parse_args()
    return args

def collocation(inp, outp, freq_filter, results, coll_type, pos):
    pos = bool(pos == 'true')
    i = str(unicode(open(inp, 'r').read(), errors='ignore'))
    o = open(outp, 'w')
    all_words = []
    if pos:
        text = i.split(' ')[:-1]
        all_words = [x[0:x.index('/')] if x != '\n' else x for x in text]
        all_words = [x.strip(' ').strip('\n') for x in all_words]

    else:
        measures = nltk.collocations.TrigramAssocMeasures()
        finder = TrigramCollocationFinder.from_words(all_words)
    finder.apply_freq_filter(int(freq_filter))
    colls = finder.nbest(measures.pmi, int(results))
    for coll in colls:
        o.write("%s\t%s" % coll)
        o.write('\n')
    o.close()

if __name__ == '__main__':
    args = Parser()

    collocation(args.input, args.output, args.freq_filter, args.results, args.coll_type, args.pos)

1:fb617586f4b2

    args = the_parser.parse_args()
    return args

def collocation(inp, outp, freq_filter, results, coll_type, pos):
    pos = bool(pos == 'true')
    with open(inp, 'r') as fd:
        i = fd.read()

    all_words = []
    if pos:
        text = i.split(' ')[:-1]
        all_words = [x[0:x.index('/')] if x != '\n' else x for x in text]
        all_words = [x.strip(' ').strip('\n') for x in all_words]

    else:
        measures = nltk.collocations.TrigramAssocMeasures()
        finder = TrigramCollocationFinder.from_words(all_words)
    finder.apply_freq_filter(int(freq_filter))
    colls = finder.nbest(measures.pmi, int(results))
    with  open(outp, 'w') as output:
        for coll in colls:
            output.write("%s\t%s" % coll)
            output.write('\n')

if __name__ == '__main__':
    args = Parser()

    collocation(args.input, args.output, args.freq_filter, args.results, args.coll_type, args.pos)