get alias by attribute of model used for validation after initialization

[Corporate-Fabian]

I have a few models with similar columns (named with a shared prefix). Later on I want to do things with those special columns. When I do this for a specific model I can identify the columns like this:

import pandera as pa
from pandera.typing import DataFrame, Series


class ExampleModelA(pa.DataFrameModel):
    a: Series[str] = pa.Field(alias="A")
    b: Series[str] = pa.Field(alias="B")
    transfer_comment: Series[str] = pa.Field(alias="Your comment")
    transfer_response: Series[str] = pa.Field(alias="Your response")


class ExampleModelB(pa.DataFrameModel):
    c: Series[str] = pa.Field(alias="C")
    d: Series[str] = pa.Field(alias="D")
    transfer_comment: Series[str] = pa.Field(alias="Value X")
    transfer_foo: Series[str] = pa.Field(alias="Value Foo")
    transfer_bar: Series[str] = pa.Field(alias="Value Bar")

# I want to replace these two lines
transfer_keys = [k for k in ExampleModelA._get_model_attrs().keys() if k.startswith("transfer")]
transfer_aliases = [ExampleModelA._get_model_attrs()[k].alias  for k in transfer_keys]

df = DataFrame[ExampleModelA]({"A": ["xxx"], "B": ["xxx"], "Your comment": ["xxx"], "Your response": ["xxx"]})

for alias in transfer_aliases:
    print(df[alias])

How, if possible, can I achieve the above but in a dynamic manner for already initialized / validated instances? Ex. in the example above, how would I do this if I only had the initialized df? (edited)

[kumare3]

is it that you want dynamically assign checks to your dataframe?

[Corporate-Fabian]

no, I want to do things "outside" pandera