質問編集履歴
1
追記
test
CHANGED
|
File without changes
|
test
CHANGED
|
@@ -34,19 +34,123 @@
|
|
|
34
34
|
|
|
35
35
|
```
|
|
36
36
|
|
|
37
|
+
from bottle import route, run
|
|
38
|
+
|
|
39
|
+
# -*- coding: utf-8 -*-
|
|
40
|
+
|
|
41
|
+
import pandas as pd
|
|
42
|
+
|
|
43
|
+
import numpy as np
|
|
44
|
+
|
|
45
|
+
import matplotlib.pyplot as plt
|
|
46
|
+
|
|
47
|
+
import json
|
|
48
|
+
|
|
49
|
+
import csv
|
|
50
|
+
|
|
51
|
+
from collections import OrderedDict
|
|
52
|
+
|
|
53
|
+
from sklearn.externals import joblib
|
|
54
|
+
|
|
55
|
+
#mecab
|
|
56
|
+
|
|
57
|
+
from natto import MeCab
|
|
58
|
+
|
|
59
|
+
nm = MeCab()
|
|
60
|
+
|
|
61
|
+
print(nm)
|
|
62
|
+
|
|
63
|
+
m = MeCab("-Owakati")
|
|
64
|
+
|
|
65
|
+
MODEL_FILE='bays_model-2.csv'
|
|
66
|
+
|
|
67
|
+
clf = joblib.load(MODEL_FILE)
|
|
68
|
+
|
|
69
|
+
|
|
70
|
+
|
|
71
|
+
#前処理
|
|
72
|
+
|
|
73
|
+
f = open('baysweb_word.csv', 'r')
|
|
74
|
+
|
|
75
|
+
line = json.load(f)
|
|
76
|
+
|
|
77
|
+
dek_dic={}
|
|
78
|
+
|
|
79
|
+
dek_dic={v:i for i, v in enumerate(line.drop("Unnamed: 0", axis=1).columns)}
|
|
80
|
+
|
|
81
|
+
def func(x):
|
|
82
|
+
|
|
83
|
+
if x == None:
|
|
84
|
+
|
|
85
|
+
return 1
|
|
86
|
+
|
|
87
|
+
else:
|
|
88
|
+
|
|
89
|
+
return 0 # NaN
|
|
90
|
+
|
|
91
|
+
|
|
92
|
+
|
|
37
93
|
@route('/housmart', method="POST")
|
|
38
94
|
|
|
39
95
|
def clean_name():
|
|
40
96
|
|
|
41
|
-
curl http://localhost:8080/housmart -X POST -H "Content-Type: application/json" -d '{"clean_buldingname": "
|
|
97
|
+
curl http://localhost:8080/housmart -X POST -H "Content-Type: application/json" -d '{"clean_buldingname": "ライオンズガーデン明大前パラダイム"}'
|
|
42
98
|
|
|
43
|
-
|
|
99
|
+
dek_dic={}
|
|
44
100
|
|
|
45
|
-
cl
|
|
101
|
+
dek_dic={v:i for i, v in enumerate(line.drop("Unnamed: 0", axis=1).columns.T)}
|
|
46
102
|
|
|
47
|
-
|
|
103
|
+
def func(x):
|
|
48
104
|
|
|
105
|
+
if x == None:
|
|
106
|
+
|
|
107
|
+
return 1
|
|
108
|
+
|
|
109
|
+
else:
|
|
110
|
+
|
|
111
|
+
return 0 # NaN
|
|
112
|
+
|
|
113
|
+
are=[]
|
|
114
|
+
|
|
115
|
+
for n in nm.parse(clean_buildingname, as_nodes=True):
|
|
116
|
+
|
|
117
|
+
if n.surface !="":
|
|
118
|
+
|
|
119
|
+
are.append(n.surface)
|
|
120
|
+
|
|
121
|
+
ree=[]
|
|
122
|
+
|
|
123
|
+
for i, v in dek_dic.items():
|
|
124
|
+
|
|
125
|
+
for n in are:
|
|
126
|
+
|
|
127
|
+
if i==n:
|
|
128
|
+
|
|
129
|
+
ree.append(v)
|
|
130
|
+
|
|
131
|
+
else:
|
|
132
|
+
|
|
133
|
+
pass
|
|
134
|
+
|
|
135
|
+
were=[]
|
|
136
|
+
|
|
137
|
+
for i,v in dek_dic.items():
|
|
138
|
+
|
|
139
|
+
if v in ree:
|
|
140
|
+
|
|
141
|
+
were.append(None)
|
|
142
|
+
|
|
143
|
+
else:
|
|
144
|
+
|
|
145
|
+
were.append(i)
|
|
146
|
+
|
|
147
|
+
xx=pd.DataFrame(were).applymap(func).T
|
|
148
|
+
|
|
149
|
+
clf.predict(xx)#予測
|
|
150
|
+
|
|
151
|
+
if clf.predict(xx)==1:
|
|
152
|
+
|
|
49
|
-
return str(clean_buildingname)
|
|
153
|
+
return str(clean_buildingname)
|
|
50
154
|
|
|
51
155
|
|
|
52
156
|
|
|
@@ -54,4 +158,6 @@
|
|
|
54
158
|
|
|
55
159
|
|
|
56
160
|
|
|
161
|
+
|
|
162
|
+
|
|
57
163
|
```
|