Start using Generic types

[paddymul]

Checks

• [x] I have checked that this enhancement has not already been requested

How would you categorize this request. You can select multiple if not sure

Developer ergonomics (defaults, error messages)

Enhancement Description

Typing DFStats, DataFlow, and BuckarooWidget are all difficult because they depend on implicit linkages

Most of DataFlow and BuckarooWidget are agnostic to the dataframe library The polars and pandas implementations of DFStats specifically exist to smooth over type incomatability

This pseudocode is a very good step towards applying generics to fix the problem while maintaining type safety

Pseudo Code Implementation

from typing_extensions import override
import pandas as pd
import polars as pl
from typing import Generic, Type, TypeVar, List

DFType = TypeVar(name='DFType')

class WorksOnDFBase(Generic[DFType]):

    @classmethod
    def get_df_columns(cls, df:DFType) -> List[str]:
        ...
class Base(Generic[DFType]):
    internal_df :DFType

    foo:Type[WorksOnDFBase[DFType]]
    
    def __init__(self, df:DFType) -> None:
      self.internal_df = df

class WorksOnPolars(WorksOnDFBase[pl.DataFrame]):
    @override
    @classmethod
    def get_df_columns(cls, df) -> List[str]:
        return [str(col) for col in df.columns]

class WorksOnPandas(WorksOnDFBase[pd.DataFrame]):
    @override
    @classmethod
    def get_df_columns(cls, df:pd.DataFrame) -> List[str]:
        return [str(col) for col in df.columns]

#this should work
class ExtendsBase(Base[pd.DataFrame]):
    foo = WorksOnPandas

#this fails because WorksOnPolars is incompatible with pd.DataFrame
class ExtendsBase2(Base[pd.DataFrame]):

    foo = WorksOnPolars

#this works work
class ExtendsBase3(Base[pl.DataFrame]):

    foo = WorksOnPolars

Prior Art

n/a

[paddymul]

Here's a better version that passes mypy

from abc import ABC, abstractmethod
from typing_extensions import override
import pandas as pd
import polars as pl
from typing import Generic, Type, TypeVar, List

DFType = TypeVar('DFType')

class WorksOnDFBase(Generic[DFType], ABC):

    @abstractmethod
    @classmethod
    def get_df_columns(cls, df:DFType) -> List[str]:
        pass

class Base(Generic[DFType]):
    internal_df :DFType

    foo:Type[WorksOnDFBase[DFType]]
    
    @abstractmethod
    def __init__(self, df:DFType) -> None:
        pass

    def get_columns(self, df:DFType) -> List[str]:
        return self.foo.get_df_columns(df)

class WorksOnPolars(WorksOnDFBase[pl.DataFrame]):
    @override
    @classmethod
    def get_df_columns(cls, df) -> List[str]:
        return [str(col) for col in df.columns]

class WorksOnPandas(WorksOnDFBase[pd.DataFrame]):
    @override
    @classmethod
    def get_df_columns(cls, df:pd.DataFrame) -> List[str]:
        return [str(col) for col in df.columns]

#this should work
class ExtendsBase(Base[pd.DataFrame]):
    foo = WorksOnPandas


# #this fails because WorksOnPolars is incompatible with pd.DataFrame
# class ExtendsBase2(Base[pd.DataFrame]):

#     foo = WorksOnPolars

#this works work
class ExtendsBase3(Base[pl.DataFrame]):

    foo = WorksOnPolars