lo-ranking/s2.py

#!/usr/bin/env python
# -*- coding:utf-8 -*-
'''
Step 2
'''
from __future__ import print_function
from base import Lo, create_csv, TARGET_DIR
import os
import csv
import sys

los = Lo()

xmodels = {
    'IMSf': {}, 'IMSD': {}, 'IMHf': {}, 'IMHD': {}, 'IFSf': {}, 'IFSD': {},
    'IFHf': {}, 'IFHD': {}, 'CMSf': {}, 'CMSD': {}, 'CMHf': {}, 'CMHD': {},
    'CFSf': {}, 'CFSD': {}, 'CFHf': {}, 'CFHD': {}, 'PMSf': {}, 'PMSD': {},
    'PMHf': {}, 'PMHD': {}, 'PFSf': {}, 'PFSD': {}, 'PFHf': {}, 'PFHD': {},
}
'''
แก้ตรงนี้ >> RANK_ORDER << ถ้าจะเพิ่ม rank ใส่อะไรก็ใส่ไปตามลำดับ
'''
RANK_ORDER = ['1', '2', '3', '4']

fs = ('Case1_LS.csv', 'Case1_Gender.csv', 'Case1_Level.csv', 'Case1_SciF.csv')


def process_s2_data():
    sum_result = {}
    rank_result = {}
    step, rank = '', ''
    for i in fs:
        f_name = '%s-result.csv' % i
        # print(f_name)
        with open(os.path.join(TARGET_DIR, f_name), 'rb') as f:
            rows = csv.reader(f)
            for r in rows:
                if r[0] in ('rank_no', 'All Result'):
                    step = r[0]
                    rank = r[1]
                    continue
                if step == 'rank_no':
                    if r[0] not in rank_result:
                        rank_result[r[0]] = {}
                    rank_result[r[0]][rank] = r[1].split(',')
                elif step == 'All Result':
                    if not r[0]:
                        continue
                    if r[0] in sum_result:
                        # print('exists:', r[1], '|', step, '|', rank)
                        pass
                    sum_result[r[0]] = r[1].split(',')
    # print(rank_result['C'].keys())
    return {
        'sum': sum_result, 'rank': rank_result
    }


def produce_s2_part1(sum_result):
    # create xxxx-1.csv
    for ii in xmodels.keys():
        rows = []
        rows.append([ii, ])
        for i in ii:
            if i not in sum_result:
                continue
            rows.append([i, ','.join(sum_result[i])])
        _f = '%s-1.csv' % ii
        create_csv(_f, rows, directory='part2')


def produce_s2_part2(rank_result):
    # create xxxx-2.csv
    for ii in xmodels.keys():
        rows = []
        for rank in RANK_ORDER:
            # print(ii, ':rank:', rank)
            rows.append(['Rank#%s' % rank, 'rank', rank])
            for i in ii:
                if i not in rank_result:
                    continue
                if rank not in rank_result[i]:
                    continue
                rows.append([i, ','.join(rank_result[i][rank])])
        _f = '%s-2.csv' % ii
        create_csv(_f, rows, directory='part2')


def process_lo_weight():
    los = {}
    for i in fs:
        f_name = '%s-output.csv' % i
        with open(os.path.join(TARGET_DIR, f_name), 'rb') as f:
            rows = csv.reader(f)
            for r in rows:
                _los = r[1].split(',')
                for _l in _los:
                    if _l not in los:
                        los[_l] = {}
                    if r[0] in los[_l]:
                        # print('exists:', _l, ':', r[0], ' > ')
                        # los[_l]['%s2' % r[0]] = r[2:]
                        pass
                    los[_l][r[0]] = r[2:]
    return los


def produce_part3(sum_result, lo_weight):
    # produce part3
    for ii in xmodels.keys():
        rows = []
        lo_xmodels = set()
        for i in ii:
            if i in sum_result:
                lo_xmodels = lo_xmodels.union(set(sum_result[i]))
        lo_xmodels = sorted(list(lo_xmodels))
        for l in lo_xmodels:
            ## build container to store sum of all weight
            lo_sum = [0 for k in xrange(0, 25)]
            if l not in lo_weight:
                continue
            for i in ii:
                if i not in lo_weight[l]:
                    continue
                lo_sum = [
                    float(x) + float(y) for x, y
                    in zip(lo_weight[l][i][9:], lo_sum)]
            rows.append([l] + lo_sum)

        _f = '%s-3.csv' % ii
        create_csv(_f, rows, directory='part2')


def main(*argv):
    _r = process_s2_data()
    _lw = process_lo_weight()
    produce_s2_part1(_r['sum'])
    produce_s2_part2(_r['rank'])
    produce_part3(_r['sum'], _lw)


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