回答編集履歴
1
pandasではなくndrrayを直接使っているので追加
test
CHANGED
|
@@ -16,9 +16,9 @@
|
|
|
16
16
|
|
|
17
17
|
0 apple $5
|
|
18
18
|
|
|
19
|
-
1 banan
|
|
19
|
+
1 banana $6
|
|
20
20
|
|
|
21
|
-
2 orag
|
|
21
|
+
2 orange $10
|
|
22
22
|
|
|
23
23
|
>>> df['price'] = df['price'].apply(lambda x: int(x.strip('$')))
|
|
24
24
|
|
|
@@ -28,8 +28,74 @@
|
|
|
28
28
|
|
|
29
29
|
0 apple 5
|
|
30
30
|
|
|
31
|
-
1 banan 6
|
|
31
|
+
1 banana 6
|
|
32
32
|
|
|
33
|
-
2 orag
|
|
33
|
+
2 orange 10
|
|
34
34
|
|
|
35
35
|
```
|
|
36
|
+
|
|
37
|
+
|
|
38
|
+
|
|
39
|
+
|
|
40
|
+
|
|
41
|
+
pandasではなくnumpyのndarrayを直接使っているようですので、ndarray版を追加します。
|
|
42
|
+
|
|
43
|
+
普通にndarrayを作ると、dtypeが'<U6'となり、整数を代入しようとしても文字列の変換されてしまいます。これを避けるためには作成時にdtypeとして'object'を指定するなどの方法をとる必要があります。
|
|
44
|
+
|
|
45
|
+
|
|
46
|
+
|
|
47
|
+
名前が縦方向の場合
|
|
48
|
+
|
|
49
|
+
```python
|
|
50
|
+
|
|
51
|
+
>>> import numpy as np
|
|
52
|
+
|
|
53
|
+
>>>
|
|
54
|
+
|
|
55
|
+
>>> fruit = np.array([['apple', '$5'], ['banana', '$6'], ['orange', '$10']])
|
|
56
|
+
|
|
57
|
+
>>> print(fruit)
|
|
58
|
+
|
|
59
|
+
[['apple' '$5']
|
|
60
|
+
|
|
61
|
+
['banana' '$6']
|
|
62
|
+
|
|
63
|
+
['orange' '$10']]
|
|
64
|
+
|
|
65
|
+
>>> fruit = np.array([[x[0], int(x[1].strip('$'))] for x in fruit], dtype='object')
|
|
66
|
+
|
|
67
|
+
>>> print(fruit)
|
|
68
|
+
|
|
69
|
+
[['apple' 5]
|
|
70
|
+
|
|
71
|
+
['banana' 6]
|
|
72
|
+
|
|
73
|
+
['orange' 10]]
|
|
74
|
+
|
|
75
|
+
```
|
|
76
|
+
|
|
77
|
+
名前が横方向の場合
|
|
78
|
+
|
|
79
|
+
```python
|
|
80
|
+
|
|
81
|
+
>>> import numpy as np
|
|
82
|
+
|
|
83
|
+
>>>
|
|
84
|
+
|
|
85
|
+
>>> fruit = np.array([['apple', 'banana', 'orange'], ['$5', '$6', '$10']])
|
|
86
|
+
|
|
87
|
+
>>> print(fruit)
|
|
88
|
+
|
|
89
|
+
[['apple' 'banana' 'orange']
|
|
90
|
+
|
|
91
|
+
['$5' '$6' '$10']]
|
|
92
|
+
|
|
93
|
+
>>> fruit = np.array([fruit[0], [int(x.strip('$')) for x in fruit[1]]], dtype = 'object')
|
|
94
|
+
|
|
95
|
+
>>> print(fruit)
|
|
96
|
+
|
|
97
|
+
[['apple' 'banana' 'orange']
|
|
98
|
+
|
|
99
|
+
[5 6 10]]
|
|
100
|
+
|
|
101
|
+
```
|