Part 2 should be completed

base.py

@@ -72,3 +72,20 @@ class Lo:
             if m:
                 r.append(m)
         return set.intersection(*[set(k) for k in r])
+
+
+def create_csv(filename, data, **kwargs):
+    if not isinstance(data, list):
+        print('Nope, data are not list')
+    directory = kwargs.get('directory', '')
+    path = os.path.join(os.getcwd(), 'build', directory)
+    try:
+        os.mkdir(path)
+    except OSError:
+        pass
+    fpath = os.path.join(path, filename)
+    with open(fpath, 'wb') as f:
+        owriter = csv.writer(
+            f, delimiter=',', quotechar='"', quoting=csv.QUOTE_ALL)
+        for r in data:
+            owriter.writerow(r)

s1.py

@@ -13,7 +13,7 @@ los = Lo()
 def get_rank_thing(filename, **kwargs):
     model_count = {
         'I': {}, 'P': {}, 'C': {}, 'M': {}, 'F': {},
-        'H': {}, 'S': {}, 'D': {}, 'F': {}
+        'H': {}, 'S': {}, 'D': {}, 'f': {}
     }
     output_filename = kwargs.get('output', '%s-output.csv' % filename)
     count = tot = lo_match = 0
@@ -48,7 +48,7 @@ def get_rank_thing(filename, **kwargs):
 def main(*argv):
     fs = ('Case1_LS.csv', 'Case1_Gender.csv',
           'Case1_Level.csv', 'Case1_SciF.csv')
-    result = ['IPC', 'MF', 'HS', 'FS']
+    result = ['IPC', 'MF', 'HS', 'fS']
     for f in xrange(0, len(fs)):
         mc = get_rank_thing(fs[f])
         x = {}

s2.py

@@ -0,0 +1,135 @@
+# -*- coding:utf-8 -*-
+#!/usr/bin/env python
+from __future__ import print_function
+from base import Lo, create_csv
+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': {},
+}
+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(os.getcwd(), 'build', f_name), 'rb') as f:
+            rows = csv.reader(f)
+            for r in rows:
+                if r[0] in ('count', 'All Result'):
+                    step = r[0]
+                    rank = r[1] if r[0] == 'count' else ''
+                    continue
+                if step == 'count':
+                    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(',')
+    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 rk in xrange(0, 3):
+            rows.append(['Rank#%s' % (rk+1), ])
+            for i in ii:
+                # find order
+                if i not in rank_result:
+                    continue
+                o = sorted(rank_result[i].keys(), reverse=True)
+                rows.append([i, ','.join(rank_result[i][o[rk]])])
+        _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(os.getcwd(), 'build', 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:
+            lo_sum = [0, 0, 0, 0, 0, 0]
+            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], 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()