add additional scalar type

dynax/custom_types.py

@@ -1,7 +1,8 @@
 import typing
-from typing import Callable, TypeAlias
+from typing import Callable, TypeAlias, Union
 
 import jaxtyping
+import numpy as np
 
 
 generating_docs = getattr(typing, "GENERATING_DOCUMENTATION", False)
@@ -13,6 +14,7 @@
 
     Scalar: TypeAlias = Array
     ScalarLike: TypeAlias = ArrayLike
+    FloatScalarLike = Union[float, Array, np.ndarray]
 elif generating_docs:
     # In the docs.
     class Scalar:
@@ -27,6 +29,8 @@ class Array:
     class ArrayLike:
         pass
 
+    FloatScalarLike = float
+
     for cls in (Scalar, ScalarLike, Array, ArrayLike):
         cls.__module__ = "builtins"
         cls.__qualname__ = cls.__name__
@@ -37,6 +41,7 @@ class ArrayLike:
 
     Scalar = jaxtyping.Shaped[Array, ""]
     ScalarLike = jaxtyping.Shaped[ArrayLike, ""]
+    FloatScalarLike = jaxtyping.Float[ArrayLike, ""]
 
 
 VectorFunc: TypeAlias = Callable[[Array], Array]

dynax/system.py

@@ -2,16 +2,16 @@
 
 from abc import abstractmethod
 from collections.abc import Callable
-from dataclasses import field, Field
-from typing import Literal, Any
+from dataclasses import Field, field
+from typing import Any, Literal
 
 import equinox
 import jax
 import jax.numpy as jnp
 import numpy as np
 from jax import Array
 
-from .custom_types import ArrayLike
+from .custom_types import FloatScalarLike
 from .util import dim2shape, pretty
 
 
@@ -165,7 +165,7 @@ def __check_init__(self):
 
     @abstractmethod
     def vector_field(
-        self, x: Array, u: Array | None = None, t: float | None = None
+        self, x: Array, u: Array | None = None, t: FloatScalarLike | None = None
     ) -> Array:
         """Compute state derivative.
 
@@ -180,7 +180,9 @@ def vector_field(
         """
         raise NotImplementedError
 
-    def output(self, x: Array, u: Array | None = None, t: float | None = None) -> Array:
+    def output(
+        self, x: Array, u: Array | None = None, t: FloatScalarLike | None = None
+    ) -> Array:
         """Compute output.
 
         Args:
@@ -203,7 +205,10 @@ def n_outputs(self) -> int | Literal["scalar"]:
         return "scalar" if y.ndim == 0 else y.shape[0]
 
     def linearize(
-        self, x0: Array | None = None, u0: Array | None = None, t: float | None = None
+        self,
+        x0: Array | None = None,
+        u0: Array | None = None,
+        t: FloatScalarLike | None = None,
     ) -> "LinearSystem":
         """Compute the Jacobian linearizationaround a point.
 
@@ -389,15 +394,17 @@ def __init__(self, sys1: AbstractSystem, sys2: AbstractSystem):
         self.n_inputs = sys1.n_inputs
 
     def vector_field(
-        self, x: Array, u: Array | None = None, t: float | None = None
+        self, x: Array, u: Array | None = None, t: FloatScalarLike | None = None
     ) -> Array:
         x1, x2 = self._unpack_states(x)
         y1 = self._sys1.output(x1, u, t)
         dx1 = self._sys1.vector_field(x1, u, t)
         dx2 = self._sys2.vector_field(x2, y1, t)
         return self._pack_states(dx1, dx2)
 
-    def output(self, x: Array, u: Array | None = None, t: float | None = None) -> Array:
+    def output(
+        self, x: Array, u: Array | None = None, t: FloatScalarLike | None = None
+    ) -> Array:
         x1, x2 = self._unpack_states(x)
         y1 = self._sys1.output(x1, u, t)
         y2 = self._sys2.output(x2, y1, t)
@@ -437,7 +444,7 @@ def __init__(self, sys1: AbstractSystem, sys2: AbstractSystem):
         self.n_inputs = sys1.n_inputs
 
     def vector_field(
-        self, x: Array, u: Array | None = None, t: float | None = None
+        self, x: Array, u: Array | None = None, t: FloatScalarLike | None = None
     ) -> Array:
         if u is None:
             u = jnp.zeros(dim2shape(self._sys1.n_inputs))
@@ -449,7 +456,9 @@ def vector_field(
         dx = self._pack_states(dx1, dx2)
         return dx
 
-    def output(self, x: Array, u: Array | None = None, t: float | None = None) -> Array:
+    def output(
+        self, x: Array, u: Array | None = None, t: FloatScalarLike | None = None
+    ) -> Array:
         x1, _ = self._unpack_states(x)
         y = self._sys1.output(x1, None, t)
         return y