-
Notifications
You must be signed in to change notification settings - Fork 3
/
Copy pathclean.py
265 lines (238 loc) · 16.9 KB
/
clean.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
import pandas as pd
import numpy as np
from sklearn.preprocessing import OneHotEncoder # encode categorical predictors
import geopandas as gpd
df = pd.read_csv('../data/gradcafe.csv')
df.loc[df.admission_via == 'POST', 'admission_via'] = 'Postal Service'
df.loc[~df.degree.isin(['PhD', 'Masters', 'Other']), 'degree'] = 'Other'
for col in ['ST', 'degree', 'admission_status', 'admission_via']:
df[col] = df[col].astype('category')
for col in ['admission_date', 'Date_added']:
df[col] = pd.to_datetime(df[col])
# scale GPA
df.Undergrad_GPA = np.where((df.Undergrad_GPA > 4) & (
df.Undergrad_GPA <= 5), df.Undergrad_GPA*4/5, df.Undergrad_GPA)
df.Undergrad_GPA = np.where((df.Undergrad_GPA > 5) & (
df.Undergrad_GPA <= 10), df.Undergrad_GPA*4/10, df.Undergrad_GPA)
# scale GRE: old to new
GRE = pd.DataFrame(list(zip(np.arange(200, 810, 10).tolist(), np.repeat(130, 4).tolist()+np.arange(131, 143).tolist() +
[143, 143, 144, 145, 146, 146, 147, 148, 149, 149, 150, 151, 151, 152, 152, 153, 154, 154, 155, 156, 156, 157, 158, 158, 159, 160, 160, 161, 162, 162, 163, 164, 164, 165, 165, 166, 167, 168, 168, 169, 169, 170, 170, 170, 170, 170])), columns=['old', 'new'])
for index, row in df.iterrows():
if df.loc[index, 'GRE_V'] > 200:
df.loc[index, 'GRE_V'] = GRE.new[GRE.old ==
df.loc[index, 'GRE_V']].to_numpy()
if df.loc[index, 'GRE_Q'] > 200:
df.loc[index, 'GRE_Q'] = GRE.new[GRE.old ==
df.loc[index, 'GRE_Q']].to_numpy()
# correct unexpected input: 179 to 169, 180 to 170
df.loc[df.GRE_V == 179, 'GRE_V'] = 169
df.loc[df.GRE_V == 180, 'GRE_V'] = 170
df.loc[df.GRE_V < 130, 'GRE_V'] = np.nan
df.loc[df.GRE_Q == 179, 'GRE_Q'] = 169
df.loc[df.GRE_Q == 180, 'GRE_Q'] = 170
df.loc[df.GRE_Q < 130, 'GRE_Q'] = np.nan
# scale GRE AW
# 0 to nan, 30 to 3, 0.3 to 3, 4.8, 42, 44 to 4
df.loc[df.GRE_W == 0, 'GRE_W'] = np.nan
df.loc[df.GRE_W == 30, 'GRE_W'] = 3
df.loc[df.GRE_W == 0.3, 'GRE_W'] = 3
df.loc[df.GRE_W == 4.8, 'GRE_W'] = 4
df.loc[df.GRE_W == 42, 'GRE_W'] = 4
df.loc[df.GRE_W == 44, 'GRE_W'] = 4
df.loc[df.GRE_W > 6, 'GRE_W'] = np.nan
# scale GRE sub
df.loc[df.GRE_sub<10, 'GRE_sub'] = np.nan
df.loc[df.GRE_sub<100, 'GRE_sub'] = df.loc[df.GRE_sub<100, 'GRE_sub']*10
# correct unexpected season
df.loc[df.season == '?', 'season'] = 'F09'
# correct name of institutions
df.loc[df.institution.str.contains(
'stanford', case=False), 'institution'] = 'Stanford University (Department of Statistics)'
df.loc[df.institution.str.contains(
'berk|UCB', case=False), 'institution'] = 'University of California--Berkeley (Department of Statistics)'
df.loc[df.institution.str.contains(
'harvard', case=False), 'institution'] = 'Harvard University (Department of Statistics)'
df.loc[df.institution.str.contains('chic', case=False) & ~df.institution.str.contains(
'loyola|illinois', case=False), 'institution'] = 'University of Chicago (Department of Statistics)'
df.loc[df.institution.str.contains(
'CMU|Mellon', case=False), 'institution'] = 'Carnegie Mellon University (Department of Statistics)'
df.loc[df.institution.str.contains('wash|seat', case=False) & ~df.institution.str.contains(
'Louis|George|state', case=False), 'institution'] = 'University of Washington (Department of Statistics)'
df.loc[df.institution.str.contains(
'duke', case=False), 'institution'] = 'Duke University (Department of Statistical Science)'
df.loc[df.institution.str.contains('Michigan', case=False) & ~df.institution.str.contains(
'State|Western|Central|Tech', case=False), 'institution'] = 'University of Michigan--Ann Arbor (Department of Statistics)'
df.loc[df.institution.str.contains('Penn|Whar', case=False) & ~df.institution.str.contains(
'state', case=False), 'institution'] = 'University of Pennsylvania (Department of Statistics)'
df.loc[df.institution.str.contains('Columbia', case=False) & ~df.institution.str.contains(
'British|Missouri|College|South', case=False), 'institution'] = 'Columbia University (Department of Statistics)'
df.loc[df.institution.str.contains('NCSU|Carolina S|NC S', case=False),
'institution'] = 'North Carolina State University--Raleigh (Department of Statistics)'
df.loc[df.institution.str.contains('Wisc|Madison|UWM', case=False) & ~df.institution.str.contains(
'Milwaukee|James', case=False), 'institution'] = 'University of Wisconsin--Madison (Department of Statistics)'
df.loc[df.institution.str.contains('North Caro|UNC|Chap', case=False) & ~df.institution.str.contains(
'State|Charlotte', case=False), 'institution'] = 'University of North Carolina--Chapel Hill (Department of Statistics & Operations Research)'
df.loc[df.institution.str.contains(
'Cornell', case=False), 'institution'] = 'Cornell University (Department of Statistical Science)'
df.loc[df.institution.str.contains(
'Iowa S|ISU', case=False), 'institution'] = 'Iowa State University (Department of Statistics)'
df.loc[df.institution.str.contains('Penn State|PSU|Pennsylvania State|Penn. State', case=False),
'institution'] = 'Pennsylvania State University (Department of Statistics)'
df.loc[df.institution.str.contains('Texas A&|TAMU|A &', case=False),
'institution'] = 'Texas A&M University--College Station (Department of Statistics)'
df.loc[df.institution.str.contains(
'UMN|nesota', case=False), 'institution'] = 'University of Minnesota--Twin Cities (School of Statistics)'
df.loc[df.institution.str.contains('Purdue', case=False) & ~df.institution.str.contains(
'India', case=False), 'institution'] = 'Purdue University--West Lafayette (Department of Statistics)'
df.loc[df.institution.str.contains('JHU|Johns|Hopkins', case=False),
'institution'] = 'Johns Hopkins University (Department of Applied Mathematics and Statistics)'
df.loc[df.institution.str.contains('Dav|daiv|Dacis', case=False),
'institution'] = 'University of California--Davis (Department of Statistics)'
df.loc[df.institution.str.contains(
'UCLA|Los', case=False), 'institution'] = 'University of California--Los Angeles (Department of Statistics)'
df.loc[df.institution.str.contains(
'Yale', case=False), 'institution'] = 'Yale University (Department of Statistics)'
df.loc[df.institution.str.contains(
'OSU|ohio s', case=False), 'institution'] = 'Ohio State University (Department of Statistics)'
df.loc[df.institution.str.contains('Illin|uiuc', case=False) & ~df.institution.str.contains(
'Chicago|North|State', case=False), 'institution'] = 'University of Illinois--Urbana-Champaign (Department of Statistics)'
df.loc[df.institution.str.contains('Rutgers', case=False) & ~df.institution.str.contains(
'education', case=False), 'institution'] = 'Rutgers University--New Brunswick (Department of Statistics and Biostatistics)'
df.loc[df.institution.str.contains('Florida|ufl', case=False) & ~df.institution.str.contains(
'central|south|State', case=False), 'institution'] = 'University of Florida (Department of Statistics)'
df.loc[df.institution.str.contains('Iowa', case=False) & ~df.institution.str.contains(
'State', case=False), 'institution'] = 'University of Iowa (Department of Statistics and Actuarial Science)'
df.loc[df.institution.str.contains(
'Rice', case=False), 'institution'] = 'Rice University (Department of Statistics)'
df.loc[df.institution.str.contains(
'Colorado St', case=False), 'institution'] = 'Colorado State University (Department of Statistics)'
df.loc[df.institution.str.contains('Florida Sta|FSU', case=False),
'institution'] = 'Florida State University (Department of Statistics)'
df.loc[df.institution.str.contains(
'conn', case=False), 'institution'] = 'University of Connecticut (Department of Statistics)'
df.loc[df.institution.str.contains('Michigan State|MSU', case=False),
'institution'] = 'University of Connecticut (Department of Statistics)'
df.loc[df.institution.str.contains(
'UCI|irvine', case=False), 'institution'] = 'University of California--Irvine (Department of Statistics)'
df.loc[df.institution.str.contains('Austin|austion', case=False),
'institution'] = 'University of Texas--Austin (Department of Statistics and Data Science)'
df.loc[df.institution.str.contains('Northwestern', case=False),
'institution'] = 'Northwestern University (Department of Statistics)'
df.loc[df.institution.str.contains(
'Pitt', case=False), 'institution'] = 'University of Pittsburgh (Department of Statistics)'
df.loc[df.institution.str.contains('george', case=False) & ~df.institution.str.contains(
'town|mason', case=False), 'institution'] = 'University of Pittsburgh (Department of Statistics)'
df.loc[df.institution.str.contains('new york|nyu', case=False) & ~df.institution.str.contains(
'state|columbia', case=False), 'institution'] = 'New York University (Department of Information, Operations, and Management Sciences)'
df.loc[df.institution.str.contains('george|uga', case=False) & ~df.institution.str.contains(
'state|tec', case=False), 'institution'] = 'University of Georgia (Department of Statistics)'
df.loc[df.institution.str.contains('Missouri', case=False) & ~df.institution.str.contains(
'S & T', case=False), 'institution'] = 'University of Missouri--Columbia (Department of Statistics)'
df.loc[df.institution.str.contains('Virginia Tech|VT|Virginia Polytechnic|VirginiaTech|Virgina Tech',
case=False), 'institution'] = 'Virginia Tech (Department of Statistics)'
df.loc[df.institution.str.contains('UCSB|Barbara|Babara|Barabara', case=False),
'institution'] = 'University of California--Santa Barbara (Department of Statistics and Applied Probability)'
df.loc[df.institution.str.contains('Indiana|IUB', case=False) & ~df.institution.str.contains(
'purdue', case=False), 'institution'] = 'Indiana University--Bloomington (Department of Statistics)'
df.loc[df.institution.str.contains('smu|Methodist', case=False),
'institution'] = 'Southern Methodist University (Department of Statistical Science)'
df.loc[df.institution.str.contains('Maryland|Baltimore', case=False) & ~df.institution.str.contains(
'college|park', case=False), 'institution'] = 'University of Maryland--Baltimore County (Department of Mathematics and Statistics)'
df.loc[df.institution.str.contains('of Virginia|Virginia University|UVA', case=False) & ~df.institution.str.contains(
'education', case=False), 'institution'] = 'University of Virginia (Department of Statistics)'
df.loc[df.institution.str.contains(
'Oregon', case=False), 'institution'] = 'Oregon State University (Department of Statistics)'
df.loc[df.institution.str.contains('UCR|Riverside', case=False),
'institution'] = 'University of California--Riverside (Department of Statistics)'
df.loc[df.institution.str.contains('Amherst|mass', case=False),
'institution'] = 'University of Massachusetts--Amherst (Department of Mathematics and Statistics)'
df.loc[df.institution.str.contains('South Carolina|Southern Carolina', case=False),
'institution'] = 'University of South Carolina (Department of Statistics)'
df.loc[df.institution.str.contains('Arizona S|ASU', case=False),
'institution'] = 'Arizona State University (School of Mathematical & Statistical Sciences)'
df.loc[df.institution.str.contains('case|Reserve|CWRU', case=False),
'institution'] = 'Case Western Reserve University (Department of Statistics)'
df.loc[df.institution.str.contains(
'temple', case=False), 'institution'] = 'Temple University (Department of Statistics)'
df.loc[df.institution.str.contains(
'baylor', case=False), 'institution'] = 'Baylor University (Department of Statistical Science)'
df.loc[df.institution.str.contains(
'Mason|GMU', case=False), 'institution'] = 'George Mason University (Department of Statistics)'
df.loc[df.institution.str.contains('Kansas S|KSU', case=False) & ~df.institution.str.contains(
'central|south|State', case=False), 'institution'] = 'Kansas State University (Department of Statistics)'
df.loc[df.institution.str.contains('Colorado Denver', case=False),
'institution'] = 'University of Colorado--Denver (Department of Mathematical and Statistical Sciences)'
df.loc[df.institution.str.contains(
'kentucky', case=False), 'institution'] = 'University of Kentucky (Department of Statistics)'
df.loc[df.institution.str.contains('Commonwealth', case=False),
'institution'] = 'Virginia Commonwealth University (Department of Statistics)'
df.loc[df.institution.str.contains(
'Diego S', case=False), 'institution'] = 'San Diego State University (Department of Mathematics and Statistics)'
df.loc[df.institution.str.contains('UNC Charlotte', case=False),
'institution'] = 'University of North Carolina--Charlotte (Department of Mathematics and Statistics)'
df.loc[df.institution.str.contains(
'Antonio', case=False), 'institution'] = 'University of Texas--San Antonio (Department of Management Science and Statistics)'
df.loc[df.institution.str.contains(
'Auburn', case=False), 'institution'] = 'Auburn University (Department of Mathematics and Statistics)'
df.loc[df.institution.str.contains('UCSD|San Diego', case=False) & ~df.institution.str.contains(
'State', case=False), 'institution'] = 'University of California--San Diego'
df.loc[df.institution.str.contains('Southern California|USC|South Califor',
case=False), 'institution'] = 'University of Southern California'
df.loc[df.institution.str.contains(
'Toronto', case=False), 'institution'] = 'University of Toronto'
df.loc[df.institution.str.contains('British Columbia', case=False) | df.institution.str.contains(
'UBC', case=False), 'institution'] = 'University of British Columbia'
df.loc[df.institution.str.contains(
'Waterl', case=False), 'institution'] = 'University of Waterloo'
# merge stat_rank and overall_rank
# add two new columns: USnew_stat_score and USnew_stat_score
df1 = pd.read_csv('../data/USnew_stat_rank.csv')
df1["postal"] = df1["district"].str.split(',', expand = True).iloc[:,1].str.strip()
df2 = pd.read_csv('../data/USnew_overall_rank.csv')
df = pd.merge(df, df1[['score', 'name', 'postal']],
left_on='institution', right_on='name')
df = df.rename(columns={'score': 'USnew_stat_score'})
df.name = df.name.str.replace(r' ?\([^)]+\)', '')
df = pd.merge(df, df2[['score', 'name']], on='name')
df = df.rename(columns={'score': 'USnew_overall_score'})
df = df.drop(columns=['name', "Unnamed: 0"])
def decision_tree_process(df):
df.loc[df.admission_status=='Wait listed', 'admission_status'] = 'Other'
df.loc[df.admission_status=='Interview', 'admission_status'] = 'Other'
df.loc[df.degree=='Other', 'degree'] = 'Masters'
df.loc[df.admission_via=='Postal Service', 'admission_via'] = 'Other'
df.loc[df.admission_via=='Phone', 'admission_via'] = 'Other'
df.loc[df.ST=='O', 'ST'] = 'I'
df = df.dropna()
X = df[['degree', 'admission_via', 'ST', 'Undergrad_GPA', 'GRE_V', 'GRE_Q', 'GRE_W', 'GRE_sub', 'USnew_stat_score', 'USnew_overall_score']]
y = df.admission_status
categorical_feature_mask = X.dtypes=='category'
ohe = OneHotEncoder(sparse=False)
X_ohe = ohe.fit_transform(X.loc[:, categorical_feature_mask])
X_ohe = pd.DataFrame(X_ohe, columns=ohe.get_feature_names(['degree', 'admission_via', 'ST']), index=X.index)
X = pd.concat([X_ohe, X.loc[:, ~categorical_feature_mask]], axis=1)
return X, y
map_us = gpd.read_file('../data/states_province/ne_110m_admin_1_states_provinces.shx')
# data munging
summary_gradcafe = df[["postal", "admission_status"]].groupby("postal").apply(lambda df: np.sum(df["admission_status"] == "Accepted") / df["admission_status"].count()).reset_index()
summary_gradcafe["admission_rate"] = df[["postal", "admission_status"]].groupby("postal").count().reset_index().iloc[:,1]
summary_gradcafe["stat_score"] = df[["USnew_stat_score", "postal"]].groupby("postal").mean().reset_index().iloc[:,1]
summary_gradcafe["overall_score"] = df[["USnew_overall_score", "postal"]].groupby("postal").mean().reset_index().iloc[:,1]
summary_gradcafe["sum_stat_score"] = df[["USnew_stat_score", "postal"]].groupby("postal").sum().reset_index().iloc[:,1]
summary_gradcafe = summary_gradcafe.rename(columns = {0 : "admission_rate", "admission_rate": "num_applicants"})
map_us = map_us.merge(summary_gradcafe, on = "postal", how = 'left')
map_us.loc[map_us["num_applicants"].isna(), "num_applicants"]=0
map_us_point = map_us.copy()
map_us_point["rep"] = map_us["geometry"].centroid
map_us_point.set_geometry("rep", inplace = True)
#make copy to set labels
map_us["stat_score2"] = map_us["stat_score"]
map_us["admission_rate2"] = map_us["admission_rate"]
map_us["overall_score2"] = map_us["overall_score"]
map_us["sum_stat_score2"] = map_us["sum_stat_score"]
map_us["num_applicants2"] = map_us["num_applicants"]
map_us.loc[map_us["sum_stat_score2"].isna(), "sum_stat_score2"] = 'NaN'
map_us.loc[map_us["overall_score2"].isna(), "overall_score2"] = 'NaN'
map_us.loc[map_us["admission_rate2"].isna(), "admission_rate2"] = 'NaN'
map_us.loc[map_us["stat_score2"].isna(), "stat_score2"] = 'NaN'
map_us.loc[map_us["num_applicants2"].isna(), "num_applicants2"] = "NaN"