FIX: set i/o types for function implementations

benchmarks/ampform.py

@@ -20,9 +20,11 @@
     from ampform.helicity import HelicityModel
     from qrules.combinatorics import StateDefinition
 
+    from tensorwaves.function import ParametrizedBackendFunction
     from tensorwaves.interface import (
         DataSample,
         FitResult,
+        Function,
         ParameterValue,
         ParametrizedFunction,
     )
@@ -55,7 +57,7 @@ def formulate_amplitude_model(
 
 def create_function(
     model: HelicityModel, backend: str, max_complexity: int | None = None
-) -> ParametrizedFunction:
+) -> ParametrizedBackendFunction:
     return create_parametrized_function(
         expression=model.expression.doit(),
         parameters=model.parameter_defaults,
@@ -66,7 +68,7 @@ def create_function(
 
 def generate_data(
     model: HelicityModel,
-    function: ParametrizedFunction,
+    function: Function[DataSample, np.ndarray],
     data_sample_size: int,
     phsp_sample_size: int,
     backend: str,
@@ -103,7 +105,7 @@ def generate_data(
 def fit(
     data: DataSample,
     phsp: DataSample,
-    function: ParametrizedFunction,
+    function: ParametrizedFunction[DataSample, np.ndarray],
     initial_parameters: Mapping[str, ParameterValue],
     backend: str,
 ) -> FitResult:

benchmarks/expression.py

@@ -61,7 +61,7 @@ def _generate_domain(
 
 def _generate_data(
     size: int,
-    function: Function,
+    function: Function[DataSample, np.ndarray],
     rng: np.random.Generator,
     bunch_size: int = 10_000,
 ) -> DataSample:

src/tensorwaves/data/__init__.py

@@ -71,7 +71,7 @@ class IntensityDistributionGenerator(DataGenerator):
     def __init__(
         self,
         domain_generator: DataGenerator,
-        function: Function,
+        function: Function[DataSample, np.ndarray],
         domain_transformer: DataTransformer | None = None,
         bunch_size: int = 50_000,
     ) -> None:

src/tensorwaves/data/transform.py

@@ -16,6 +16,7 @@
 from ._attrs import to_tuple
 
 if TYPE_CHECKING:  # pragma: no cover
+    import numpy as np
     import sympy as sp
 
 
@@ -55,7 +56,9 @@ def __call__(self, data: DataSample) -> DataSample:
 class SympyDataTransformer(DataTransformer):
     """Implementation of a `.DataTransformer`."""
 
-    def __init__(self, functions: Mapping[str, Function]) -> None:
+    def __init__(
+        self, functions: Mapping[str, Function[DataSample, np.ndarray]]
+    ) -> None:
         if any(not isinstance(f, Function) for f in functions.values()):
             msg = (
                 f"Not all values in the mapping are an instance of {Function.__name__}"
@@ -64,7 +67,7 @@ def __init__(self, functions: Mapping[str, Function]) -> None:
         self.__functions = dict(functions)
 
     @property
-    def functions(self) -> dict[str, Function]:
+    def functions(self) -> dict[str, Function[DataSample, np.ndarray]]:
         """Read-only access to the internal mapping of functions."""
         return dict(self.__functions)
 

src/tensorwaves/estimator.py

@@ -29,7 +29,7 @@ def create_cached_function(
     backend: str,
     free_parameters: Iterable[sp.Symbol],
     use_cse: bool = True,
-) -> tuple[ParametrizedFunction, DataTransformer]:
+) -> tuple[ParametrizedFunction[DataSample, np.ndarray], DataTransformer]:
     """Create a function and data transformer for cached computations.
 
     Once it is known which parameters in an expression are to be optimized, this
@@ -118,7 +118,7 @@ class ChiSquared(Estimator):
 
     def __init__(  # noqa: PLR0913
         self,
-        function: ParametrizedFunction,
+        function: ParametrizedFunction[DataSample, np.ndarray],
         domain: DataSample,
         observed_values: np.ndarray,
         weights: np.ndarray | None = None,
@@ -185,7 +185,7 @@ class UnbinnedNLL(Estimator):
 
     def __init__(  # noqa: PLR0913
         self,
-        function: ParametrizedFunction,
+        function: ParametrizedFunction[DataSample, np.ndarray],
         data: DataSample,
         phsp: DataSample,
         phsp_volume: float = 1.0,

src/tensorwaves/function/__init__.py

@@ -3,9 +3,10 @@
 from __future__ import annotations
 
 import inspect
-from typing import TYPE_CHECKING, Callable, Iterable, Mapping
+from typing import Callable, Iterable, Mapping
 
 import attrs
+import numpy as np
 from attrs import field, frozen
 
 from tensorwaves.interface import (
@@ -15,9 +16,6 @@
     ParametrizedFunction,
 )
 
-if TYPE_CHECKING:
-    import numpy as np
-
 
 def _all_str(
     _: PositionalArgumentFunction, __: attrs.Attribute, value: Iterable[str]
@@ -66,7 +64,7 @@ def _to_tuple(argument_order: Iterable[str]) -> tuple[str, ...]:
 
 
 @frozen
-class PositionalArgumentFunction(Function):
+class PositionalArgumentFunction(Function[DataSample, np.ndarray]):
     """Wrapper around a function with positional arguments.
 
     This class provides a :meth:`~.Function.__call__` that can take a `.DataSample` for
@@ -90,7 +88,7 @@ def __call__(self, data: DataSample) -> np.ndarray:
         return self.function(*args)
 
 
-class ParametrizedBackendFunction(ParametrizedFunction):
+class ParametrizedBackendFunction(ParametrizedFunction[DataSample, np.ndarray]):
     """Implements `.ParametrizedFunction` for a specific computational back-end.
 
     .. seealso:: :func:`.create_parametrized_function`

src/tensorwaves/interface.py

@@ -41,7 +41,7 @@ def __call__(self, data: InputType) -> OutputType: ...
 """Allowed types for parameter values."""
 
 
-class ParametrizedFunction(Function[DataSample, np.ndarray]):
+class ParametrizedFunction(Function[InputType, OutputType]):
     """Interface of a callable function.
 
     A `ParametrizedFunction` identifies certain variables in a mathematical expression

tests/optimizer/test_fit_simple_model.py

@@ -38,7 +38,7 @@ def generate_domain(
 def generate_data(
     size: int,
     boundaries: dict[str, tuple[float, float]],
-    function: Function,
+    function: Function[DataSample, np.ndarray],
     rng: np.random.Generator,
     bunch_size: int = 10_000,
 ) -> DataSample: