BUG: DataFrame.select_dtypes returns reference to original DataFrame

[villebro]

Pandas version checks

• I have checked that this issue has not already been reported.

• I have confirmed this bug exists on the latest version of pandas.

• I have confirmed this bug exists on the main branch of pandas.

Reproducible Example

from decimal import Decimal
import pandas as pd


df = pd.DataFrame({"a": [1, 2, None, 4], "b": [1.5, None, 2.5, 3.5]})
selected_df = df.select_dtypes(include=["number", Decimal])
selected_df.fillna(0, inplace=True)

df.to_markdown()

Issue Description

On Pandas 1.3.4 and previously, mutating the DataFrame returned by select_dtypes didn't change the original DataFrame. However, on 1.4.3, it appears mutating the selected DataFrame changes the original one. I suspect this is the PR that has caused the change in behavior: #42611

On 1.4.3, the original DataFrame is mutated:

	a	b
0	1	1.5
1	2	0
2	0	2.5
3	4	3.5

Expected Behavior

on 1.3.4, the original DataFrame is unchanged:

	a	b
0	1	1.5
1	2	nan
2	nan	2.5
3	4	3.5

Installed Versions

INSTALLED VERSIONS ------------------ commit : e8093ba python : 3.8.13.final.0 python-bits : 64 OS : Darwin OS-release : 21.6.0 Version : Darwin Kernel Version 21.6.0: Sat Jun 18 17:07:22 PDT 2022; root:xnu-8020.140.41~1/RELEASE_ARM64_T6000 machine : arm64 processor : arm byteorder : little LC_ALL : None LANG : None LOCALE : None.UTF-8

pandas : 1.4.3
numpy : 1.23.2
pytz : 2022.2.1
dateutil : 2.8.2
setuptools : 63.2.0
pip : 22.2.2
Cython : None
pytest : None
hypothesis : None
sphinx : None
blosc : None
feather : None
xlsxwriter : None
lxml.etree : 4.9.1
html5lib : None
pymysql : None
psycopg2 : None
jinja2 : 3.1.2
IPython : 8.4.0
pandas_datareader: None
bs4 : 4.11.1
bottleneck : None
brotli : None
fastparquet : None
fsspec : None
gcsfs : None
markupsafe : 2.1.1
matplotlib : None
numba : None
numexpr : None
odfpy : None
openpyxl : None
pandas_gbq : None
pyarrow : None
pyreadstat : None
pyxlsb : None
s3fs : None
scipy : None
snappy : None
sqlalchemy : None
tables : None
tabulate : 0.8.10
xarray : None
xlrd : None
xlwt : None
zstandard : None

[jorisvandenbossche]

@villebro thanks for the report! That's indeed an unintended change in behaviour in 1.4 (and #42611 seems indeed the likely cause). I opened a PR to fix this: #48176

[villebro]

@jorisvandenbossche thanks for looking into this issue and fixing it! FWIW, PR reviewed and approved 👍

[simonjayhawkins]

That's indeed an unintended change in behaviour in 1.4 (and #42611 seems indeed the likely cause)

can confirm, from bisect, first bad commit: [37ea2cd] PERF: select_dtypes (#42611)

[pickfire]

I noticed this broke our code in 1.5 since we are depending on the select_dtypes to return a reference to the original data frame in 1.4 as a feature, so that we can do .select_dtypes(exclude=['int8']).fillna('', inplace=True), I tried finding this but can't seemed to find it in the list of API breaking change.

https://pandas.pydata.org/pandas-docs/stable/whatsnew/v1.5.0.html#backwards-incompatible-api-changes

Can we also put there?