質問編集履歴
1
写真じゃなくコードを書くこと
test
CHANGED
|
File without changes
|
test
CHANGED
|
@@ -1,6 +1,12 @@
|
|
|
1
1
|
string型をint型に直したところ、
|
|
2
2
|
|
|
3
3
|
欠損値がなかったのに、欠損値が生成されてしまいました。
|
|
4
|
+
|
|
5
|
+
最初のdftest_train[preds_name].isnull().any()ではFalseなのに、
|
|
6
|
+
|
|
7
|
+
データ成形後
|
|
8
|
+
|
|
9
|
+
次のdftest_train[preds_name].isnull().any()ではTrueになっている（欠損値あり）
|
|
4
10
|
|
|
5
11
|
なぜなのか、教えて頂けると幸いです。
|
|
6
12
|
|
|
@@ -8,8 +14,118 @@
|
|
|
8
14
|
|
|
9
15
|
|
|
10
16
|
|
|
11
|
-
|
|
17
|
+
```python
|
|
12
18
|
|
|
13
|
-
|
|
19
|
+
print(df.shape)
|
|
14
20
|
|
|
21
|
+
print(dftest.shape)
|
|
22
|
+
|
|
23
|
+
print(df_train.shape)
|
|
24
|
+
|
|
25
|
+
print(dftest_train.shape)
|
|
26
|
+
|
|
27
|
+
print(df_train[preds_name].isnull().any())
|
|
28
|
+
|
|
29
|
+
print(dftest_train[preds_name].isnull().any())
|
|
30
|
+
|
|
31
|
+
```
|
|
32
|
+
|
|
33
|
+
(6461, 157)
|
|
34
|
+
|
|
35
|
+
(4273, 157)
|
|
36
|
+
|
|
37
|
+
(6461, 157)
|
|
38
|
+
|
|
39
|
+
(4272, 157)
|
|
40
|
+
|
|
41
|
+
False
|
|
42
|
+
|
|
43
|
+
False
|
|
44
|
+
|
|
45
|
+
```python
|
|
46
|
+
|
|
47
|
+
# 間違っているセル（2019/06/26）
|
|
48
|
+
|
|
49
|
+
if len(df_train.loc[:, :preds_name]) != 0:
|
|
50
|
+
|
|
51
|
+
# df_train用
|
|
52
|
+
|
|
53
|
+
# object型の要素をint, float型に変換する
|
|
54
|
+
|
|
55
|
+
# object型の要素を探す
|
|
56
|
+
|
|
57
|
+
factorize = []
|
|
58
|
+
|
|
59
|
+
for i, j in zip(df_train.columns, df_train.dtypes):
|
|
60
|
+
|
|
61
|
+
factorize.append(i)
|
|
62
|
+
|
|
63
|
+
# 0~要素数の種類の数を割り当てる
|
|
64
|
+
|
|
65
|
+
for i in factorize:
|
|
66
|
+
|
|
67
|
+
id_1, id_2 = df_train[i].factorize()
|
|
68
|
+
|
|
69
|
+
id_3 = pd.DataFrame({i: id_1})
|
|
70
|
+
|
|
71
|
+
df_train[i] = id_3
|
|
72
|
+
|
|
73
|
+
del factorize
|
|
74
|
+
|
|
75
|
+
|
|
76
|
+
|
|
77
|
+
if len(dftest_train.loc[:, :preds_name]) != 0:
|
|
78
|
+
|
|
79
|
+
# dftest_train用
|
|
80
|
+
|
|
81
|
+
# object型の要素をint, float型に変換する
|
|
82
|
+
|
|
83
|
+
# object型の要素を探す
|
|
84
|
+
|
|
85
|
+
factorize = []
|
|
86
|
+
|
|
87
|
+
for i, j in zip(dftest_train.columns, dftest_train.dtypes):
|
|
88
|
+
|
|
89
|
+
factorize.append(i)
|
|
90
|
+
|
|
91
|
+
# 0~要素数の種類の数を割り当てる
|
|
92
|
+
|
|
93
|
+
for i in factorize:
|
|
94
|
+
|
|
15
|
-
|
|
95
|
+
id_1, id_2 = dftest_train[i].factorize()
|
|
96
|
+
|
|
97
|
+
id_3 = pd.DataFrame({i: id_1})
|
|
98
|
+
|
|
99
|
+
dftest_train[i] = id_3
|
|
100
|
+
|
|
101
|
+
del factorize
|
|
102
|
+
|
|
103
|
+
print(df_train[preds_name].isnull().any())
|
|
104
|
+
|
|
105
|
+
print(dftest_train[preds_name].isnull().any())
|
|
106
|
+
|
|
107
|
+
print(len(df_train))
|
|
108
|
+
|
|
109
|
+
print(len(dftest_train))
|
|
110
|
+
|
|
111
|
+
print(len(df_test))
|
|
112
|
+
|
|
113
|
+
print(len(dftest_test))
|
|
114
|
+
|
|
115
|
+
print(df_train.loc[0, preds_name])
|
|
116
|
+
|
|
117
|
+
```
|
|
118
|
+
|
|
119
|
+
False
|
|
120
|
+
|
|
121
|
+
True
|
|
122
|
+
|
|
123
|
+
6461
|
|
124
|
+
|
|
125
|
+
4272
|
|
126
|
+
|
|
127
|
+
0
|
|
128
|
+
|
|
129
|
+
1
|
|
130
|
+
|
|
131
|
+
0
|