feat: skbase inheritance in central base classes (#10)

Implements #8 - without additional testing - by changing inheritance to
`BaseObject` in the base bootstrapper and the base bootstrap config.

Main changes in this refactor:

* bootstrap classes now inherit from `scikit-base` `BaseObject`
* methods provided by `BaseObject` are removed - `__repr__`, `__str__`,
`__hash__`
* methods provided by `BaseObject` but not currently present are added -
`get_params`, `set_params`

Depends on #23, as
`poetry.lock` will lock the dependencies in place (so imo it should be
removed).

---------

Co-authored-by: Sankalp Gilda <[email protected]>

Author: fkiraly

pyproject.toml

@@ -23,6 +23,7 @@ homepage = "https://tsbootstrap.readthedocs.io/en/latest/"
 
 [tool.poetry.dependencies]
 python = ">=3.10,<3.12"
+scikit-base = ">= 0.6.1"
 arch = "~5.6"
 hmmlearn = "~0.3"
 pyclustering = "~0.10"

src/tsbootstrap/base_bootstrap.py

@@ -8,6 +8,7 @@
 
 import numpy as np
 from scipy.stats import rv_continuous
+from skbase.base import BaseObject
 
 from tsbootstrap.utils.odds_and_ends import time_series_split
 
@@ -24,7 +25,7 @@
 from tsbootstrap.tsfit import TSFitBestLag
 
 
-class BaseTimeSeriesBootstrap(metaclass=ABCMeta):
+class BaseTimeSeriesBootstrap(BaseObject):
     """
     Base class for time series bootstrapping.
 
@@ -34,6 +35,8 @@ class BaseTimeSeriesBootstrap(metaclass=ABCMeta):
         If n_bootstraps is not greater than 0.
     """
 
+    _tags = {"object_type": "bootstrap"}
+
     def __init__(self, config: BaseTimeSeriesBootstrapConfig) -> None:
         self.config = config
 
@@ -131,14 +134,14 @@ def _generate_samples(
             else:
                 yield data
 
-    @abstractmethod
     def _generate_samples_single_bootstrap(
         self, X: np.ndarray, y: np.ndarray | None = None
     ) -> tuple[list[np.ndarray], list[np.ndarray]]:
         """Generates list of bootstrapped indices and samples for a single bootstrap iteration.
 
         Should be implemented in derived classes.
         """
+        raise NotImplementedError("abstract method")
 
     def _check_input(self, X):
         """Checks if the input is valid."""
@@ -154,24 +157,6 @@ def get_n_bootstraps(
         """Returns the number of bootstrapping iterations."""
         return self.config.n_bootstraps  # type: ignore
 
-    def __repr__(self) -> str:
-        """Returns the string representation of the object."""
-        return f"{self.__class__.__name__}(config={self.config})"
-
-    def __str__(self) -> str:
-        """Returns the string representation of the object."""
-        return f"{self.__class__.__name__}(config={self.config})"
-
-    def __eq__(self, __value: object) -> bool:
-        """Returns True if the objects are equal, False otherwise."""
-        if not isinstance(__value, BaseTimeSeriesBootstrap):
-            return NotImplemented
-        return self.config == __value.config
-
-    def __hash__(self) -> int:
-        """Returns the hash of the object."""
-        return hash(self.config)
-
 
 class BaseResidualBootstrap(BaseTimeSeriesBootstrap):
     """
@@ -234,24 +219,6 @@ def _fit_model(self, X: np.ndarray, y: np.ndarray | None = None) -> None:
             self.order = fit_obj.get_order()
             self.coefs = fit_obj.get_coefs()
 
-    def __repr__(self) -> str:
-        """Returns the string representation of the object."""
-        return f"{self.__class__.__name__}(config={self.config})"
-
-    def __str__(self) -> str:
-        """Returns the string representation of the object."""
-        return f"{self.__class__.__name__}(config={self.config})"
-
-    def __eq__(self, __value: object) -> bool:
-        """Returns True if the objects are equal, False otherwise."""
-        if not isinstance(__value, BaseResidualBootstrap):
-            return NotImplemented
-        return self.config == __value.config
-
-    def __hash__(self) -> int:
-        """Returns the hash of the object."""
-        return hash((super().__hash__(), self.config))
-
 
 class BaseMarkovBootstrap(BaseResidualBootstrap):
     """
@@ -288,24 +255,6 @@ def __init__(
 
         self.hmm_object = None
 
-    def __repr__(self) -> str:
-        """Returns the string representation of the object."""
-        return f"{self.__class__.__name__}(config={self.config})"
-
-    def __str__(self) -> str:
-        """Returns the string representation of the object."""
-        return self.__repr__()
-
-    def __eq__(self, __value: object) -> bool:
-        """Returns True if the objects are equal, False otherwise."""
-        if not isinstance(__value, BaseMarkovBootstrap):
-            return NotImplemented
-        return self.config == __value.config
-
-    def __hash__(self) -> int:
-        """Returns the hash of the object."""
-        return hash((super().__hash__(), self.config))
-
 
 class BaseStatisticPreservingBootstrap(BaseTimeSeriesBootstrap):
     """Bootstrap class that generates bootstrapped samples preserving a specific statistic.
@@ -352,24 +301,6 @@ def _calculate_statistic(self, X: np.ndarray) -> np.ndarray:
         statistic_X = self.config.statistic(X, **kwargs_stat)
         return statistic_X
 
-    def __repr__(self) -> str:
-        """Returns the string representation of the object."""
-        return f"{self.__class__.__name__}(config={self.config})"
-
-    def __str__(self) -> str:
-        """Returns the string representation of the object."""
-        return self.__repr__()
-
-    def __eq__(self, __value: object) -> bool:
-        """Returns True if the objects are equal, False otherwise."""
-        if not isinstance(__value, BaseStatisticPreservingBootstrap):
-            return NotImplemented
-        return self.config == __value.config
-
-    def __hash__(self) -> int:
-        """Returns the hash of the object."""
-        return hash((super().__hash__(), self.config))
-
 
 # We can only fit uni-variate distributions, so X must be a 1D array, and `model_type` in BaseResidualBootstrap must not be "var".
 class BaseDistributionBootstrap(BaseResidualBootstrap):
@@ -450,24 +381,6 @@ def _fit_distribution(
         resids_dist_params = resids_dist.fit(resids)
         return resids_dist, resids_dist_params
 
-    def __repr__(self) -> str:
-        """Returns the string representation of the object."""
-        return f"{self.__class__.__name__}(config={self.config})"
-
-    def __str__(self) -> str:
-        """Returns the string representation of the object."""
-        return self.__repr__()
-
-    def __eq__(self, __value: object) -> bool:
-        """Returns True if the objects are equal, False otherwise."""
-        if not isinstance(__value, BaseDistributionBootstrap):
-            return NotImplemented
-        return self.config == __value.config
-
-    def __hash__(self) -> int:
-        """Returns the hash of the object."""
-        return hash((super().__hash__(), self.config))
-
 
 class BaseSieveBootstrap(BaseResidualBootstrap):
     """
@@ -536,21 +449,3 @@ def _fit_resids_model(self, X: np.ndarray) -> None:
             self.resids_fit_model = resids_fit_model
             self.resids_order = resids_order
             self.resids_coefs = resids_coefs
-
-    def __repr__(self) -> str:
-        """Returns the string representation of the object."""
-        return f"{self.__class__.__name__}(config={self.config})"
-
-    def __str__(self) -> str:
-        """Returns the string representation of the object."""
-        return self.__repr__()
-
-    def __eq__(self, __value: object) -> bool:
-        """Returns True if the objects are equal, False otherwise."""
-        if not isinstance(__value, BaseSieveBootstrap):
-            return NotImplemented
-        return self.config == __value.config
-
-    def __hash__(self) -> int:
-        """Returns the hash of the object."""
-        return hash((super().__hash__(), self.config))