Lie group variables
Built-in variable types for optimization on Lie groups.
-
class jaxls.SE2Var[source]
SE2Var(id: ‘jax.Array | int’)
-
retract_fn(delta: ndarray | Array | Any) → MatrixLieGroup | Any
Manifold right plus. Computes T’ = T @ exp(delta).
Supports pytrees containing Lie group instances recursively; simple Euclidean
addition will be performed for all other arrays.
- Parameters:
-
- Return type:
MatrixLieGroup | Any
-
tangent_dim: ClassVar[int] = 3
Dimension of the tangent space.
-
class jaxls.SE3Var[source]
SE3Var(id: ‘jax.Array | int’)
-
retract_fn(delta: ndarray | Array | Any) → MatrixLieGroup | Any
Manifold right plus. Computes T’ = T @ exp(delta).
Supports pytrees containing Lie group instances recursively; simple Euclidean
addition will be performed for all other arrays.
- Parameters:
-
- Return type:
MatrixLieGroup | Any
-
tangent_dim: ClassVar[int] = 6
Dimension of the tangent space.
-
class jaxls.SO2Var[source]
SO2Var(id: ‘jax.Array | int’)
-
retract_fn(delta: ndarray | Array | Any) → MatrixLieGroup | Any
Manifold right plus. Computes T’ = T @ exp(delta).
Supports pytrees containing Lie group instances recursively; simple Euclidean
addition will be performed for all other arrays.
- Parameters:
-
- Return type:
MatrixLieGroup | Any
-
tangent_dim: ClassVar[int] = 1
Dimension of the tangent space.
-
class jaxls.SO3Var[source]
SO3Var(id: ‘jax.Array | int’)
-
retract_fn(delta: ndarray | Array | Any) → MatrixLieGroup | Any
Manifold right plus. Computes T’ = T @ exp(delta).
Supports pytrees containing Lie group instances recursively; simple Euclidean
addition will be performed for all other arrays.
- Parameters:
-
- Return type:
MatrixLieGroup | Any
-
tangent_dim: ClassVar[int] = 3
Dimension of the tangent space.