QST: Replace subset of DataFrame by another subset of DataFrame

[zxdawn]

• I have searched the [pandas] tag on StackOverflow for similar questions.

• I have asked my usage related question on StackOverflow.

---

Question about pandas

Note: If you'd still like to submit a question, please read this guide detailing how to provide the necessary information for us to reproduce your question.

# Your code here, if applicable
import numpy as np
import pandas as pd
dic = {'A': [1, 4, 1, 4], 'B': [9, 2, 5, 3], 'C': [0, 0, 5, 3]}
df = pd.DataFrame(dic)
df.index = [0, 0, 1, 1]

df1 = df.copy()
df1[:] = np.nan

df1.loc[0]['A'] = df.loc[0]['A']

original df:

   A  B  C
0  1  9  0
0  4  2  0
1  1  5  5
1  4  3  3

I want to replace the subset of the copy with the subset of original df:

    A   B   C
0 1   NaN NaN
0 4   NaN NaN
1 NaN NaN NaN
1 NaN NaN NaN

However, it shows no change at all:

    A   B   C
0 NaN NaN NaN
0 NaN NaN NaN
1 NaN NaN NaN
1 NaN NaN NaN

[MarcoGorelli]

Hi @zxdawn - please ask usage questions on StackOverflow first

Did you not get a SettingWithCopyWarning when executing df1.loc[0]['A'] = df.loc[0]['A']?

Use

df1.loc[0, 'A'] = 

for assignment

[zxdawn]

Thanks, @MarcoGorelli. It works now.

BTW, I didn't get the warning with pandas = 1.0.5.

[MarcoGorelli]

Thanks, @MarcoGorelli. It works now.

BTW, I didn't get the warning with pandas = 1.0.5.

Thanks for the report - I can confirm that 1.0.5 doesn't give this warning with this snippet, while 1.1.2 does. Reopening for now with needs tests, will check later when this was fixed

[MarcoGorelli]

760ba37a776947694ec816eaf29ce8937e08f544 is the first bad commit
commit 760ba37a776947694ec816eaf29ce8937e08f544
Author: jbrockmendel <jbrockmendel@gmail.com>
Date:   Tue May 26 15:47:18 2020 -0700

    CLN: _consolidate_inplace less (#34389)

 pandas/core/generic.py            |  9 ---------
 pandas/core/internals/managers.py | 10 ----------
 2 files changed, 19 deletions(-)
bisect run success

[samilAyoub]

take