skrobot.sdf.SignedDistanceFunction

class skrobot.sdf.SignedDistanceFunction(origin, coords=None, use_abs=False)

A base class for signed distance functions (SDFs).

Suffixes _obj and _sdf (e.g. in points_sdf) indicate that points are expressed in the sdf’s object coordinate (self.coords) and sdf-specfic coordinate respectively. Each SDF performs the signed-distance computation in its own sdf-specific coordinate. For example, the origin of GridSDF’s sdf-specific coordinate is the tip of the precomputed-gridded-box.

Usually, a child class implements the computation in the sdf-specific coordinates and SignedDistanceFunction wraps them so that the user can pass and get points and values expressed in an object coordinate. Thus, it is less likely that a user directly calls a method in a child class.

Methods

__call__(points_obj)

Compute signed distances of input points to the implicit surface.

Parameters:

points_obj (numpy.ndarray[float](n_point, 3)) – 2 dim point array w.r.t. object coordinate.

Returns:

signed_distances – 1 dim (n_point,) array of signed distance.

Return type:

numpy.ndarray[float]

on_surface(points_obj)

Check if points are on the surface.

Parameters:

points_obj (numpy.ndarray[float](n_point, 3)) – 2 dim point array w.r.t. an object coordinate.

Returns:

• logicals (numpy.ndarray[bool](n_point,)) – boolean vector of the on-surface predicate for points_obj.

• sd_vals (numpy.ndarray[float](n_point,)) – signed distances corresponding to logicals.

surface_points(n_sample=1000)

Sample points from the implicit surface of the sdf.

Parameters:

n_sample (int) – number of sample points.

Returns:

• points_obj (numpy.ndarray[float](n_point, 3)) – sampled points w.r.t object coordinate.

• dists (numpy.ndarray[float](n_point,)) – signed distances corresponding to points_obj.

__eq__(value, /)

Return self==value.

__ne__(value, /)

Return self!=value.

__lt__(value, /)

Return self<value.

__le__(value, /)

Return self<=value.

__gt__(value, /)

Return self>value.

__ge__(value, /)

Return self>=value.