lo-ranking/s1.py

#!/usr/bin/env python
# -*- coding:utf-8 -*-
'''
Step 1
'''
from __future__ import print_function
from base import Lo, RAW_DIR, TARGET_DIR
import os
import csv
import sys


los = Lo()


def get_rank_thing(filename, **kwargs):
    model_count = {
        'I': {}, 'P': {}, 'C': {}, 'M': {}, 'F': {},
        'H': {}, 'S': {}, 'D': {}, 'f': {}
    }
    output_filename = kwargs.get('output', '%s-output.csv' % filename)
    count = tot = lo_match = 0
    with open(os.path.join(TARGET_DIR, output_filename), 'wb') as o:
        owriter = csv.writer(o, delimiter=',', quotechar='"',
                             quoting=csv.QUOTE_ALL)
        with open(os.path.join(RAW_DIR, filename)) as f:
            reader = csv.reader(f)
            reader.next()
            for r in reader:
                mn, rank_no = r[0], r[9]
                w = [
                    r[1], r[2], r[3], r[4], r[5], r[6], r[7], r[8],  # July 1
                    rank_no,  # July 1
                ]
                ## add 24 model's weight
                for i in xrange(10, 34):
                    w.append(r[i])
                q = (
                    ('A', r[1]), ('B', r[2]), ('C', r[3]), ('D', r[4]),
                    ('E', r[5]), ('F', r[6]), ('G', r[7]), ('H', r[8]),
                )
                result = los.get_mm_tuple(q)
                tot += 1
                lo_match += len(result)
                if result:
                    # print(u'%s > %s' % (r[0], list(result)))
                    count += 1
                    owriter.writerow([
                        mn,
                        ','.join(list(result)),
                        len(result),
                        # rank_no
                    ] + w)
                    if rank_no not in model_count[mn]:
                        model_count[mn][rank_no] = set()
                    model_count[mn][rank_no] = model_count[mn][rank_no].union(result)
    print(u'[%s]\n    Count = %s | Total = %s | Lo_w_match = %s' % (
        filename, count, tot, lo_match))
    return model_count


def main(*argv):
    fs = ('Case1_LS.csv', 'Case1_Gender.csv',
          'Case1_Level.csv', 'Case1_SciF.csv')
    result = ['IPC', 'MF', 'HS', 'fD']
    for f in xrange(0, len(fs)):
        mc = get_rank_thing(fs[f])
        x = {}
        for i in result[f]:
            x[i] = mc[i]
        # print('x: ', x)
        mx = [set(j) for i, j in x.items()]
        order = sorted(list(set.union(*mx)))
        sum_file = '%s-result.csv' % fs[f]
        with open(os.path.join(TARGET_DIR, sum_file), 'wb') as o:
            result_writer = csv.writer(o, delimiter=',', quotechar='"',
                                       quoting=csv.QUOTE_ALL)
            m = {}
            # print(j, ':', order)
            for i in xrange(0, len(order)):
                result_writer.writerow(['rank_no', order[i]])
                # print('count , ', order[i], ' / ', type(order[i]))
                for j in x.keys():
                    try:
                        # print(j, ', ', s)
                        lo_result = mc[j][order[i]]
                        # print(order[i], ' == ', len(lo_result))
                        # if i == 2 and len(mc[j]) > 3:
                        #     for _i in xrange(3, len(mc[j])):
                        #         lo_result = lo_result.union(mc[j][order[_i]])
                        #         # print(' ==EX== ', len(lo_result))
                        if j not in m:
                            m[j] = lo_result
                        else:
                            m[j] = m[j].union(lo_result)
                        result_writer.writerow([j, ','.join(list(lo_result))])
                    except KeyError:
                        print(' no key ', j, '[', order[i], ']')
            result_writer.writerow(['All Result', ''])
            for _m in m.keys():
                result_writer.writerow([_m, ','.join(list(m[_m]))])

if __name__ == '__main__':
    if len(sys.argv) > 1:
        main(sys.argv[1:])
    else:
        main()