check a nexus file's site patterns

check_nexus.py

#!/usr/bin/env python
# -*- coding: utf-8 -*-

"""
(c) 2014 Brant Faircloth || http://faircloth-lab.org/
All rights reserved.

This code is distributed under a 3-clause BSD license. Please see
LICENSE.txt for more information.

Created on 16 July 2014 13:56 PDT (-0700)
"""

import re
import numpy
from collections import Counter

all = []
with open("Q30-QD2-MISS_0.75.recode.snapp.nexus", "rU") as infile:
    for line in infile:
        if line.startswith("wesj"):
            result = re.search(".*\s+(.*)\n", line)
            all.append(list(result.groups()[0]))
a = numpy.asarray(all)
# slice by column
characters = 0
with_missing = 0
no_missing = 0
cnt = Counter()
for colnum in xrange(a.shape[1]):
    characters += 1
    col = a[:, colnum]
    colset = set(col)
    if "?" in colset:
        with_missing += 1
    if "?" not in colset:
        no_missing += 1
    cnt[" ".join(list(colset))] += 1
print "characters", characters
print "sites with missing", with_missing
print "sites with no missing", no_missing
print cnt