用另一个 Dataframe 的值填充 Dataframe（不是相同的列名）

Question

我正在尝试用另一个 dataframe (InputData) 中的值填充 Python 中的空 dataframe (OutputData)。

InputData 有四列（“Strike”、“DTE”、“IV”、“Pred_IV”）OutputData 具有来自 InputData 的所有唯一 Strikes 作为索引，并且作为列名称，所有来自 Input Data 的唯一 DTE。

我的目标是用来自 InputData 的相应“Pred_IV”值填充 OutputData。 因为它需要同时匹配索引和列名，所以我没有考虑如何使用任何已知的 function 进行匹配。

如果 InputData 中没有与索引和列名匹配的值，则该值可以保持为 NaN

在下面找到我使用 df.to_dict() 提取的数据帧以获取更多详细信息。

非常感谢您的帮助。

最好的，弗洛

输入数据.head()

    Strike  DTE     IV      Pred_IV
8   0.5131  2.784   0.3366  0.733360
9   0.5131  3.781   0.3291  0.735295
20  0.5864  2.784   0.3178  0.733476
21  0.5864  3.781   0.3129  0.735357
22  0.5864  4.778   0.3008  0.736143

InputData.head().to_dict()

{'Strike': {8: 0.5131, 9: 0.5131, 20: 0.5864, 21: 0.5864, 22: 0.5864},
 'DTE': {8: 2.784, 9: 3.781, 20: 2.784, 21: 3.781, 22: 4.778},
 'IV': {8: 0.33659999999999995,
  9: 0.32909999999999995,
  20: 0.3178,
  21: 0.3129,
  22: 0.30079999999999996},
 'Pred_IV': {8: 0.7333602770095773,
  9: 0.7352946387206533,
  20: 0.7334762408944806,
  21: 0.7353567361456718,
  22: 0.7361431377881676}})

输出数据.head()

        0.025   0.101   0.197   0.274   0.523   0.772   1.769   2.267   2.784   3.781   4.778   5.774
0.5131  NaN     NaN     NaN     NaN     NaN     NaN     NaN     NaN     NaN     NaN     NaN     NaN
0.5864  NaN     NaN     NaN     NaN     NaN     NaN     NaN     NaN     NaN     NaN     NaN     NaN
0.6597  NaN     NaN     NaN     NaN     NaN     NaN     NaN     NaN     NaN     NaN     NaN     NaN
0.7330  NaN     NaN     NaN     NaN     NaN     NaN     NaN     NaN     NaN     NaN     NaN     NaN
0.7697  NaN     NaN     NaN     NaN     NaN     NaN     NaN     NaN     NaN     NaN     NaN     NaN

OutputData.head(2).to_dict()

{0.025: {0.5131: nan,
  0.5864: nan,
  0.6597: nan,
  0.733: nan,
  0.7696999999999999: nan},
 0.101: {0.5131: nan,
  0.5864: nan,
  0.6597: nan,
  0.733: nan,
  0.7696999999999999: nan},
 0.197: {0.5131: nan,
  0.5864: nan,
  0.6597: nan,
  0.733: nan,
  0.7696999999999999: nan},
 0.274: {0.5131: nan,
  0.5864: nan,
  0.6597: nan,
  0.733: nan,
  0.7696999999999999: nan},
 0.523: {0.5131: nan,
  0.5864: nan,
  0.6597: nan,
  0.733: nan,
  0.7696999999999999: nan},
 0.772: {0.5131: nan,
  0.5864: nan,
  0.6597: nan,
  0.733: nan,
  0.7696999999999999: nan},
 1.769: {0.5131: nan,
  0.5864: nan,
  0.6597: nan,
  0.733: nan,
  0.7696999999999999: nan},
 2.267: {0.5131: nan,
  0.5864: nan,
  0.6597: nan,
  0.733: nan,
  0.7696999999999999: nan},
 2.784: {0.5131: nan,
  0.5864: nan,
  0.6597: nan,
  0.733: nan,
  0.7696999999999999: nan},
 3.781: {0.5131: nan,
  0.5864: nan,
  0.6597: nan,
  0.733: nan,
  0.7696999999999999: nan},
 4.778: {0.5131: nan,
  0.5864: nan,
  0.6597: nan,
  0.733: nan,
  0.7696999999999999: nan},
 5.774: {0.5131: nan,
  0.5864: nan,
  0.6597: nan,
  0.733: nan,
  0.7696999999999999: nan}}

Answer 1

这是一种方法来做我认为你的问题是问：

import pandas as pd
import numpy as np
InputData = pd.DataFrame(
    columns='Strike,DTE,IV,Pred_IV'.split(','),
    index=[8,9,20,21,22],
    data=[[0.5131,  2.784,   0.3366,  0.733360],
    [0.5131,  3.781,   0.3291,  0.735295],
    [0.5864,  2.784,   0.3178,  0.733476],
    [0.5864,  3.781,   0.3129,  0.735357],
    [0.5864,  4.778,   0.3008,  0.736143]])

OutputData = pd.DataFrame(data=np.NaN,
    columns=pd.Index(name='DTE', data=list(set(InputData.DTE.to_list()))), 
    index=pd.Index(name='Strike', data=list(set(InputData.Strike.to_list()))))
def foo(x):
    OutputData.loc[x.Strike, x.DTE] = x.Pred_IV
InputData.apply(foo, axis=1)
print(OutputData)

Output：

DTE        2.784     3.781     4.778
Strike
0.5131  0.733360  0.735295       NaN
0.5864  0.733476  0.735357  0.736143

如果你更喜欢未命名的索引，你可以这样做：

OutputData = pd.DataFrame(data=np.NaN,
    columns=list(set(InputData.DTE.to_list())),
    index=list(set(InputData.Strike.to_list())))

Output：

           2.784     3.781     4.778
0.5131  0.733360  0.735295       NaN
0.5864  0.733476  0.735357  0.736143