correct fairness testing

tests/test_general/bar_pass_data_preparing.py

@@ -0,0 +1,36 @@
+import pandas as pd
+
+
+# auxiliary functions for data preparing
+def grouper_race(x):
+    if x == 7:
+        return 1
+    else: 
+        return 0
+
+
+def grouper_gpa(x):
+    if x > 3.4:
+        return 2
+    elif x < 3.1:
+        return 0
+    else:
+        return 1
+
+
+def prepare_data(name='bar_pass_prediction'):
+
+    if name == 'bar_pass_prediction':
+        d = pd.read_csv('bar_pass_prediction.csv')
+        for x in ['ID', 'race1', 'race2', 'sex', 'bar', 'dnn_bar_pass_prediction', 'pass_bar', 'indxgrp2',
+                 'gender', 'grad', 'Dropout', 'fulltime', 'lsat', 'zfygpa', 'ugpa', 'zgpa', 'other', 'asian',
+                 'black', 'hisp']:
+            del d[x]
+
+        d['gpa'] = d['gpa'].apply(grouper_gpa)
+        d['race'] = d['race'].apply(grouper_race)
+        d = d.rename(columns = {'gpa': 'target', 'race': 'attr'})
+        d = pd.get_dummies(d, drop_first=True)
+        d = d.dropna(how='any')
+
+        return d