BUG: arguments dtype and parse_dates of read_csv does not work properly

[godslayer-inferno]

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

import pandas as pd
df = pd.DataFrame([{'name': 'foo','date': pd.Timestamp('2024-01-23T00:00:00Z')},
                   {'name': 'bar','date': pd.Timestamp('2024-01-23T12:00:00Z')}])
df.to_csv('test.csv', index=False)

df_read = pd.read_csv('test.csv', dtype=str, parse_dates=['date'])
print(df_read.dtypes) # The dtype of column 'date' should be datetime-like, but we got 'object' instead

Issue Description

Reading a csv file and trying to specify both dtype and parse_dates will not parse the columns specified in parse_dates as dates, but strings instead

Expected Behavior

The code in the example should print:
name object
date datetime64[ns, UTC]
dtype: object

Installed Versions

INSTALLED VERSIONS

commit : f538741
python : 3.11.5.final.0
python-bits : 64
OS : Windows
OS-release : 10
Version : 10.0.22621
machine : AMD64
processor : Intel64 Family 6 Model 183 Stepping 1, GenuineIntel
byteorder : little
LC_ALL : None
LANG : None
LOCALE : Chinese (Simplified)_China.936

pandas : 2.2.0
numpy : 1.24.3
pytz : 2023.3.post1
dateutil : 2.8.2
setuptools : 68.0.0
pip : 23.3
Cython : None
pytest : 7.4.0
hypothesis : None
sphinx : None
blosc : None
feather : None
xlsxwriter : None
lxml.etree : 4.9.3
html5lib : 1.1
pymysql : 1.0.2
psycopg2 : None
jinja2 : 3.1.2
IPython : 8.15.0
pandas_datareader : None
adbc-driver-postgresql: None
adbc-driver-sqlite : None
bs4 : 4.12.2
bottleneck : None
dataframe-api-compat : None
fastparquet : None
fsspec : 2023.9.2
gcsfs : None
matplotlib : 3.7.2
numba : 0.58.0
numexpr : 2.8.7
odfpy : None
openpyxl : None
pandas_gbq : None
pyarrow : 11.0.0
pyreadstat : None
python-calamine : None
pyxlsb : None
s3fs : None
scipy : 1.11.3
sqlalchemy : 2.0.22
tables : None
tabulate : 0.9.0
xarray : None
xlrd : 2.0.1
zstandard : None
tzdata : 2023.3
qtpy : None
pyqt5 : None

[aidoskanapyanov]

I confirm this on main branch

>>> import pandas as pd
>>> print(pd.__version__)
3.0.0.dev0+163.g7281475054.dirty
>>> df = pd.DataFrame([{'name': 'foo','date': pd.Timestamp('2024-01-23T00:00:00Z')},
        ...                    {'name': 'bar','date': pd.Timestamp('2024-01-23T12:00:00Z')}])
>>> df.to_csv('test.csv', index=False)
>>> df_read = pd.read_csv('test.csv', dtype=str, parse_dates=['date'])
>>> print(df_read.dtypes)
name    object
date    object
dtype: object

[MarcoGorelli]

this looks like it's as-documented, you set a dtype for the entire dataframe and it was respected

[MarcoGorelli]

you can call to_datetime after having read the csv file

closing then but thanks for the report

[philpep]

this looks like it's as-documented, you set a dtype for the entire dataframe and it was respected

Sure but parse_dates was ignored. Doesn't this deserve at least a warning or a changelog notice ?

The behavior has changed between 2.1.4 and 2.2.0, also parsing the date seem different (datetime vs date ?)

import io

import pandas as pd

df = pd.read_csv(
    io.StringIO("date,value\n2020-09-01,3\n"), dtype=object, parse_dates=["date"]
)
print(df)
print(df.dtypes)

With pandas 2.1.4:

        date value
0 2020-09-01     3
date     datetime64[ns]
value            object
dtype: object

With pandas 2.2.0:

                  date value
0  2020-09-01 00:00:00     3
date     object
value    object
dtype: object

[MarcoGorelli]

ah hadn't realised it was a regression, sorry - reopening then, running git bisect now

[MarcoGorelli]

this changed in #56047

[MarcoGorelli]

I think this is correct to be honest, closing again then