Custom Check Methods don't support custom error (any more)

[harryjordancma]

Describe the bug
I'm trying to pass a custom error message to a custom check that I have however I get an error that the error message has multiple arguments. This was previous possible before this commit #1574.

• I have checked that this issue has not already been reported.
• I have confirmed this bug exists on the latest version of pandera.
• (optional) I have confirmed this bug exists on the main branch of pandera.

Code Sample, a copy-pastable example

import pandera as pa
import pandera.extensions as extensions
import pandas as pd



@extensions.register_check_method(statistics=["min_value", "max_value"])
def is_between(pandas_obj, *, min_value, max_value):
    return (min_value <= pandas_obj) & (pandas_obj <= max_value)


schema = pa.DataFrameSchema({
    "col": pa.Column(int, pa.Check.is_between(min_value=1, max_value=10, error="Value not in between", raise_warning =True))
})

data = pd.DataFrame({"col": [1, 5, 11]})
schema.validate(data)

This produces this error message

/home/ubuntu/miniconda/envs/py311/lib/python3.11/site-packages/pandera/backends/pandas/base.py:140: SchemaWarning: Column 'col' failed element-wise validator number 0: This is small. failure cases: 11
  warnings.warn(
---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
Cell In[5], line 14
      7 schema = pa.DataFrameSchema({
      8     "col": pa.Column(int, pa.Check.less_than(10, error="This is small.", raise_warning = True))
      9 })
     11 schema.validate(data)
     13 schema2 = pa.DataFrameSchema({
---> 14     "col": pa.Column(int, pa.Check.is_between(min_value=1, max_value=10, error="This is small, but positive.", raise_warning = True))
     15 })
     18 schema2.validate(data)

File ~/miniconda/envs/py311/lib/python3.11/site-packages/pandera/api/extensions.py:130, in register_check_statistics.<locals>.register_check_statistics_decorator.<locals>._wrapper(cls, *args, **kwargs)
    128     arg_names = statistics_args
    129 args_dict = {**dict(zip(arg_names, args)), **kwargs}
--> 130 check = class_method(cls, *args, **kwargs)
    131 check.statistics = {
    132     stat: args_dict.get(stat) for stat in statistics_args
    133 }
    134 check.statistics_args = statistics_args

File ~/miniconda/envs/py311/lib/python3.11/site-packages/pandera/api/extensions.py:294, in register_check_method.<locals>.register_check_wrapper.<locals>.check_method(cls, *args, **kwargs)
    291 error_stats = ", ".join(f"{k}={v}" for k, v in stats.items())
    292 error = f"{check_fn.__name__}({error_stats})" if stats else None
--> 294 return cls(
    295     partial(check_fn_wrapper, **stats),
    296     name=check_fn.__name__,
    297     error=error,
    298     **validate_check_kwargs(check_kwargs),
    299 )

TypeError: pandera.api.checks.Check() got multiple values for keyword argument 'error'

Expected behavior

Expected the check to raise the custom error message as this is the standard behavior in the built in checks https://pandera.readthedocs.io/en/stable/reference/generated/pandera.api.checks.Check.html

I'd really like to have this be the same with custom checks as I'd like the checks to be easier to understand.

Desktop (please complete the following information):

• OS: Ubuntu
• Browser: N/A
• Version: 0.19.3

[verwindle]

Reproduced the bug on every version since 0.19.0b2. This example works from 0.19.0b1 and below without any issues.
And 0.19.0b2 requires pyspark to be installed.

Explicit error=... at line 297 here (

pandera/pandera/api/extensions.py

Line 297 in d2bfed0

error=error,

) is leading to this error. Keyword arguments error and check_kwargs overlap. The error parameter is computed inside check_method like its hardcoded. But as I see, it should fall to this value if only the error parameter is not present in kwargs. Otherwise, it can't be changed.