主页 / user-30101162

Jian Wang's questions

Martin Hope
Jian Wang
Asked: 2025-03-29 16:59:47 +0800 CST
  • 6

我正在尝试分配一个带有嵌套列表的数据框单元格：

df.loc['y','A'] = [[2]]

但实际分配的值是[2]。

它适用于预期[2], [[[2]]], [[[[2]]]]，但不适用于[[2]]

请参阅以下代码：

import pandas as pd
import numpy as np

df = pd.DataFrame({"A": [[1], [[2]], [[[3]]], [[[[4]]]], np.array([[2]]), np.array([[[2]]]), [[1],[2]]], 
                   "B": [[1], [[2]], [[[3]]], [[[[4]]]], np.array([[2]]), np.array([[[2]]]), [[1],[2]]],
                   "C": [1,2,3,4,5,6,7]
                   }, 
                   index=["x", "y", "z", "w","a","b","c"])


# initial assing works
print(df)

df.loc['x','A'] = [1] # good
df.loc['y','A'] = [[2]] # buggy, actual assigned value [2]
df.loc['z','A'] = [[[3]]] # good
df.loc['w','A'] = [[[[4]]]] #good

df.loc['a','A'] = np.array([[2]], dtype=object) # buggy, actual assign value [2]
df.loc['b','A'] = np.array([[[2]]], dtype=object) # good


#df.loc['b','A'] = [1,2] # error: Must have equal len keys and value when setting with an iterable
df.loc['c','A'] = [[1],[2]] # buggy, actual assigned value [1,2]

print(df)

输出：

            A           B  C
x         [1]         [1]  1
y       [[2]]       [[2]]  2
z     [[[3]]]     [[[3]]]  3
w   [[[[4]]]]   [[[[4]]]]  4
a       [[2]]       [[2]]  5
b     [[[2]]]     [[[2]]]  6
c  [[1], [2]]  [[1], [2]]  7
           A           B  C
x        [1]         [1]  1
y        [2]       [[2]]  2
z    [[[3]]]     [[[3]]]  3
w  [[[[4]]]]   [[[[4]]]]  4
a        [2]       [[2]]  5
b    [[[2]]]     [[[2]]]  6
c     [1, 2]  [[1], [2]]  7

更奇怪的是，如果我们删除 col "C" ，上面的所有代码注释中都不会再有 bug，也不会有错误。

arrays