回答編集履歴
1
Update
test
CHANGED
|
@@ -1,3 +1,63 @@
|
|
|
1
|
+
> csvファイルの1~3段目ごとに2次元配列としてnumpy変換したいです
|
|
2
|
+
|
|
3
|
+
|
|
4
|
+
|
|
5
|
+
行ごとに要素数が異なりますので、`numpy.ndarray` のリストに変換するのはどうでしょうか。
|
|
6
|
+
|
|
7
|
+
|
|
8
|
+
|
|
9
|
+
```python
|
|
10
|
+
|
|
11
|
+
import numpy as np
|
|
12
|
+
|
|
13
|
+
from ast import literal_eval
|
|
14
|
+
|
|
15
|
+
|
|
16
|
+
|
|
17
|
+
filename = 'data.csv'
|
|
18
|
+
|
|
19
|
+
|
|
20
|
+
|
|
21
|
+
with open(filename, 'r') as f:
|
|
22
|
+
|
|
23
|
+
data = [np.array(literal_eval(l.replace('"', '')))*10 for l in f]
|
|
24
|
+
|
|
25
|
+
|
|
26
|
+
|
|
27
|
+
from pprint import pprint
|
|
28
|
+
|
|
29
|
+
pprint(data)
|
|
30
|
+
|
|
31
|
+
|
|
32
|
+
|
|
33
|
+
#
|
|
34
|
+
|
|
35
|
+
[array([[1020, 1420],
|
|
36
|
+
|
|
37
|
+
[3970, 1240]]),
|
|
38
|
+
|
|
39
|
+
array([[2100, 750],
|
|
40
|
+
|
|
41
|
+
[2060, 2740],
|
|
42
|
+
|
|
43
|
+
[2450, 4260]]),
|
|
44
|
+
|
|
45
|
+
array([[3220, 1940],
|
|
46
|
+
|
|
47
|
+
[2970, 2930],
|
|
48
|
+
|
|
49
|
+
[2460, 3730],
|
|
50
|
+
|
|
51
|
+
[1620, 4330]])]
|
|
52
|
+
|
|
53
|
+
```
|
|
54
|
+
|
|
55
|
+
|
|
56
|
+
|
|
57
|
+
**以前の回答**
|
|
58
|
+
|
|
59
|
+
|
|
60
|
+
|
|
1
61
|
`ast.literal_eval` を使ってもよいかと思います。
|
|
2
62
|
|
|
3
63
|
```python
|