Why does @check_types overwrites custom dataframe attributes?

[SebBanDev]

Following an idea from stackoverflow to better keep track of data frames in logging messages
( https://stackoverflow.com/a/31727504/7269299 )

I assigned a .name attribute to some of my data frames in my code.
It seems that @check_types removes custom attributes from data frames:

import pandas as pd
import pandera as pa
from pandera import check_types
from pandera.typing import DataFrame


class Model(pa.SchemaModel):
    """ Example Model """


@check_types
def dummy_func(data:  DataFrame[Model]):
    print(data.name)


data = pd.DataFrame()
data.name = 'some name'
print(data.name)
dummy_func(data)

This will raise an error as .name won't be available anymore inside the dumm_func whereas when I comment out @check_types this code works fine.
Is this expected behavior? What can I do not to lose the attribute?

[sandeepkokare]

@SebbanSms correct , i can see the below error after assignment a custom variable to DF
File "C:\Users\kosa01\Anaconda3\envs\ppq\lib\site-packages\pandas\core\generic.py", line 5274, in getattr
return object.getattribute(self, name)
AttributeError: 'DataFrame' object has no attribute 'name'

we need workaround or a solution to fix this kind of error

[jeffzi]

Your .name is lost because pandera copies the DataFrame by default. pandas.DataFrame.copy() does not seem to preserve "external" attributes.

import pandas as pd

data = pd.DataFrame()
data.name = "some name"
data = data.copy()
print(data.name)
#> Traceback (most recent call last):
#> /tmp/ipykernel_79911/3401063958.py in <module>
#> ----> 1 print(data.name)
#> ~/.pyenv/versions/3.8.11/lib/python3.8/site-packages/pandas/core/generic.py in __getattr__(self, name)
#>    5485         ):
#>    5486             return self[name]
#> -> 5487         return object.__getattribute__(self, name)
#>    5488 
#>    5489     def __setattr__(self, name: str, value) -> None:
#> AttributeError: 'DataFrame' object has no attribute 'name'

You can disable the copy with @check_types(inplace=True).

[sandeepkokare]

this leads to SettingWithCopyWarning warning for DF

[jeffzi]

@sandeepkokare Can you provide a reproducible example? I don't see a warning when using @check_types(inplace=True) on @SebbanSms's example.

[SebBanDev]

Oh I also get the SettingWithCopyWarning using inplace=True in my code

[...]\pandera\schemas.py:347: SettingWithCopyWarning: 
A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
  obj[matched_colname] = _try_coercion(

[jeffzi]

Do you get a warning with the example in your initial post? If not, can you provide an example of code triggering the warning? Also, what pandera version did you install?

We can convert this discussion into an issue if we can confirm and reproduce the warning.

[SebBanDev]

I tried to reproduce it with a simple example but I as not able to 😬 let me try to investigate how to reproduce this :D