adam.pytorch.torch_like

adam.pytorch.torch_like#

Classes#

`TorchLike`	Class wrapping pyTorch types
`TorchLikeFactory`	Abstract class for a generic Array wrapper. Every method should be implemented for every data type.
`SpatialMath`	Class implementing the main geometric functions used for computing rigid-body algorithm

Module Contents#

class adam.pytorch.torch_like.TorchLike[source]#

Bases: adam.core.spatial_math.ArrayLike

Class wrapping pyTorch types

array: torch.Tensor[source]#

__post_init__()[source]#: Converts array to the desired type

__setitem__(idx, value: TorchLike | numpy.typing.ArrayLike) → TorchLike[source]#: Overrides set item operator

__getitem__(idx)[source]#: Overrides get item operator

property shape[source]#

reshape(*args)[source]#

property T: TorchLike[source]#: Returns: TorchLike: transpose of array

__matmul__(other: TorchLike | numpy.typing.ArrayLike) → TorchLike[source]#: Overrides @ operator

__rmatmul__(other: TorchLike | numpy.typing.ArrayLike) → TorchLike[source]#: Overrides @ operator

__mul__(other: TorchLike | numpy.typing.ArrayLike) → TorchLike[source]#: Overrides * operator

__rmul__(other: TorchLike | numpy.typing.ArrayLike) → TorchLike[source]#: Overrides * operator

__truediv__(other: TorchLike | numpy.typing.ArrayLike) → TorchLike[source]#: Overrides / operator

__add__(other: TorchLike | numpy.typing.ArrayLike) → TorchLike[source]#: Overrides + operator

__radd__(other: TorchLike | numpy.typing.ArrayLike) → TorchLike[source]#: Overrides + operator

__sub__(other: TorchLike | numpy.typing.ArrayLike) → TorchLike[source]#: Overrides - operator

__rsub__(other: TorchLike | numpy.typing.ArrayLike) → TorchLike[source]#: Overrides - operator

__neg__() → TorchLike[source]#: Overrides - operator

class adam.pytorch.torch_like.TorchLikeFactory[source]#

Bases: adam.core.spatial_math.ArrayLikeFactory

Abstract class for a generic Array wrapper. Every method should be implemented for every data type.

static zeros(*x: int) → TorchLike[source]#

Returns:: zero matrix of dimension *x
Return type:: TorchLike

static eye(x: int) → TorchLike[source]#

Parameters:: x (int) – dimension
Returns:: identity matrix of dimension x
Return type:: TorchLike

static array(x: numpy.typing.ArrayLike) → TorchLike[source]#

Returns:: vector wrapping x
Return type:: TorchLike

class adam.pytorch.torch_like.SpatialMath[source]#

Bases: SpatialMath

Class implementing the main geometric functions used for computing rigid-body algorithm

Parameters:: ArrayLike – abstract class describing a generic Array wrapper. It needs to be implemented for every data type

static sin(x: numpy.typing.ArrayLike) → TorchLike[source]#

Parameters:: x (ntp.ArrayLike) – angle value
Returns:: sin value of x
Return type:: TorchLike

static cos(x: numpy.typing.ArrayLike) → TorchLike[source]#

Parameters:: x (ntp.ArrayLike) – angle value
Returns:: cos value of x
Return type:: TorchLike

static outer(x: numpy.typing.ArrayLike, y: numpy.typing.ArrayLike) → TorchLike[source]#

Parameters:

x (ntp.ArrayLike) – vector
y (ntp.ArrayLike) – vector

Returns:

outer product of x and y

Return type:

TorchLike

static skew(x: TorchLike | numpy.typing.ArrayLike) → TorchLike[source]#

Parameters:: x (Union[TorchLike, ntp.ArrayLike]) – vector
Returns:: skew matrix from x
Return type:: TorchLike

static vertcat(*x: numpy.typing.ArrayLike) → TorchLike[source]#

Returns:: vertical concatenation of x
Return type:: TorchLike

static horzcat(*x: numpy.typing.ArrayLike) → TorchLike[source]#

Returns:: horizontal concatenation of x
Return type:: TorchLike