-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathmodel1_gp.py
166 lines (137 loc) · 5.68 KB
/
model1_gp.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
#!/usr/bin/env python2
# -*- coding: utf-8 -*-
"""
Created on Fri Oct 13 14:54:41 2017
@author: ldong
"""
import gc
import numpy as np
import pandas as pd
from sklearn.preprocessing import StandardScaler, LabelEncoder
def ProjectOnMedian(data1, data2, columnName):
grpOutcomes = data1.groupby(list([columnName]))['y'].median().reset_index()
grpCount = data1.groupby(list([columnName]))['y'].count().reset_index()
grpOutcomes['cnt'] = grpCount.y
grpOutcomes.drop('cnt', inplace=True, axis=1)
outcomes = data2['y'].values
x = pd.merge(data2[[columnName, 'y']], grpOutcomes,
suffixes=('x_', ''),
how='left',
on=list([columnName]),
left_index=True)['y']
return x.values
directory = '../../../input/'
tr = pd.read_csv(directory+'train_2016_v2.csv', low_memory=False)
tr17 = pd.read_csv(directory+'train_2017.csv', low_memory=False)
samp = pd.read_csv(directory+'sample_submission.csv', low_memory=False)
prop = pd.read_csv(directory+'properties_2016.csv', low_memory=False)
prop17 = pd.read_csv(directory+'properties_2017.csv', low_memory=False)
## make sure parcelid order
#sub = pd.read_csv('../../../input/sample_submission.csv')
#sub.rename(columns={'ParcelId':'parcelid'}, inplace=True)
#sub_pid = list(sub.parcelid)
#properties.set_index('parcelid')
#properties.reindex(sub_pid)
#properties.loc[:,'parcelid'] = sub.parcelid
#properties17.set_index('parcelid')
#properties17.reindex(sub_pid)
#properties17.loc[:,'parcelid'] = sub.parcelid
#%%
def data_proc(train, properties, sample):
properties.hashottuborspa = properties.hashottuborspa.astype(str)
properties.fireplaceflag = properties.fireplaceflag.astype(str)
for c in properties.columns:
if properties[c].dtype == 'object':
print(c)
lbl = LabelEncoder()
lbl.fit(list(properties[c].values))
properties[c] = lbl.transform(list(properties[c].values))
highcardinality = ['airconditioningtypeid',
'architecturalstyletypeid',
'buildingclasstypeid',
'buildingqualitytypeid',
'decktypeid',
'fips',
'hashottuborspa',
'heatingorsystemtypeid',
'pooltypeid10',
'pooltypeid2',
'pooltypeid7',
'propertycountylandusecode',
'propertylandusetypeid',
'regionidcity',
'regionidcounty',
'regionidneighborhood',
'regionidzip',
'storytypeid',
'typeconstructiontypeid',
'fireplaceflag',
'taxdelinquencyflag']
sample = sample.rename(columns={'ParcelId':'parcelid'})
sample.head()
train = train.merge(properties, how='left', on='parcelid')
test = sample.merge(properties, how='left', on='parcelid')
train['month'] = pd.DatetimeIndex(train['transactiondate']).month
test['month'] = -1
logerrors = train.logerror.ravel()
train.drop(['logerror','transactiondate'],inplace=True,axis=1)
test = test[train.columns]
train.insert(1,'nans',train.isnull().sum(axis=1))
test.insert(1,'nans',train.isnull().sum(axis=1))
train['y'] = logerrors
test['y'] = np.nan
from sklearn.model_selection import KFold
blindloodata = None
folds = 20
kf = KFold(n_splits=folds,shuffle=True,random_state=42)
for i, (train_index, test_index) in enumerate(kf.split(range(train.shape[0]))):
print('Fold:',i)
blindtrain = train.loc[test_index].copy()
vistrain = train.loc[train_index].copy()
for c in highcardinality:
blindtrain.insert(1,'loo'+c, ProjectOnMedian(vistrain,
blindtrain,c))
if(blindloodata is None):
blindloodata = blindtrain.copy()
else:
blindloodata = pd.concat([blindloodata,blindtrain])
for c in highcardinality:
test.insert(1,'loo'+c, ProjectOnMedian(train,
test,c))
test.drop(highcardinality,inplace=True,axis=1)
train = blindloodata
train.drop(highcardinality,inplace=True,axis=1)
feats = train.columns[1:-1]
for c in feats:
train[c] = train[c].fillna(train[c].median())
test[c] = test[c].fillna(train[c].median())
return train, test
#%%
train, test = data_proc(tr, prop, samp)
train17, test17 = data_proc(tr17, prop17, samp)
train17.loc[:,'month'] = train17.loc[:,'month']+12
train = pd.concat([train,train17], axis=0).reset_index(drop=True)
test = test17
xtrain = train.loc[(train.y>-0.418)&(train.y< 0.418)].copy()
xtrain = xtrain.reset_index(drop=True)
features = xtrain.columns[1:-1]
xtrain.y = xtrain.y+.418
xtrain.y /= (2*.418)
valid = xtrain.loc[(train.month>=19)&(train.month<=21)]
valid.y = valid.y*2*0.418-0.418
xtrain = xtrain.loc[train.month<=18]
ss = StandardScaler()
ss.fit(pd.concat([xtrain[features],test[features]]))
xtrain[features] = ss.transform(xtrain[features])
valid[features] = ss.transform(valid[features])
test[features] = ss.transform(test[features])
from gp_feat import *
from sklearn.metrics import mean_absolute_error
#validpreds = GP(valid)
#print 'valid mae: %.6f'%mean_absolute_error(validpreds,valid.y)
testpreds = GP(test)
#%%
sub = pd.read_csv(directory+'sample_submission.csv')
for i, c in enumerate(sub.columns[sub.columns != 'ParcelId']):
sub[c] = testpreds
sub.to_csv('gp17.csv',index=False,float_format='%.4f')