lo-ranking/s8.py

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


COMPARSION_PAIR = [
    (0, 0), (1, 1), (4, 2), (5, 3),
    (8, 4), (10, 5), (11, 6), (13, 7),
    (15, 8), (16, 9), (17, 10), (19, 11),
    (20, 12), (22, 13), (25, 14),
    (27, 15), (28, 16), (29, 17), (31, 18),
]


def get_test_data(m):
    if not m:
        return []
    result = []
    lo_list = []
    fpath = os.path.join(RAW_DIR, 'LO_TestData.csv')
    with open(fpath, 'rb') as f:
        rows = csv.reader(f)
        rows.next()
        for r in rows:
            if r[1] != m.upper():
                continue
            result.append(r)
            lo_list.append(r[0])
    return (lo_list, result)


def get_ro_data(m, suffix):
    if not m or not suffix:
        return []
    result = []
    fname = '%s-%s.csv' % (m, suffix)
    fpath = os.path.join(TARGET_DIR, 'part7', fname)
    with open(fpath, 'rb') as f:
        rows = csv.reader(f)
        for r in rows:
            result.append(r)
    return result


def get_similarity_count(l, r, **kwargs):
    C_PAIR = kwargs.get('pair', COMPARSION_PAIR)
    cnt = 0
    for (x, y) in C_PAIR:
        r_value = r[y].replace('*', '')
        if l[x] == r_value:
            cnt += 1
        elif r_value == 'Y/N' and l[x] in ('Y', 'N'):
            cnt += 1
    return cnt


def middleman(ls, rs):
    if not len(ls):
        return ls
    result = []
    for lo in ls:
        data = lo[:]
        for ro in rs:
            sim_count = get_similarity_count(lo[:], ro)
            data.append(sim_count)
        result.append(data)
    return result


def main():
    header = ['LO_xxx', '____', ]
    for m in MODEL_LIST:
        l_list, lo_data = get_test_data(m)
        for i in '1234':
            m_header = header[:]
            m_header[1] = m
            ros = get_ro_data(m, i)
            m_header += [r[0] for r in ros]
            result = middleman(lo_data, ros)
            output_name = '%s-%s.csv' % (m, i)
            # process header -- get ind of first R_xxx
            number_ind = 0
            cnt_ind = 0
            for i in result[0]:
                try:
                    float(i)
                    number_ind = cnt_ind
                    cnt_ind = 0
                    break
                except ValueError:
                    cnt_ind += 1
            __h = m_header[:2] + ['_' for i in xrange(0, number_ind-2)] + m_header[2:]
            result = [__h] + result
            create_csv(output_name, result, directory='part8')


if __name__ == '__main__':
    main()