CircularDetector.surface_deriv¶

CircularDetector.surface_deriv(self, param)[source]¶

Return the surface derivative at param.

The derivative at parameter phi is given by

deriv = R * radius * (-sin(phi), -cos(phi))

where R is a rotation matrix.

Parameters

paramfloat or array-like: Parameter value(s) at which to evaluate.

Returns

derivnumpy.ndarray: Array representing the derivative vector(s) at param. If param is a single parameter, the returned array has shape (2,), otherwise param.shape + (2,).

See also

surface

Examples

The method works with a single parameter, resulting in a single vector:

>>> part = odl.uniform_partition(-np.pi / 2, np.pi / 2, 10)
>>> det = CircularDetector(part, axis=[1, 0], radius=2)
>>> det.surface_deriv(0)
array([ 2.,  0.])

It is also vectorized, i.e., it can be called with multiple parameters at once (or an n-dimensional array of parameters):

>>> np.round(det.surface_deriv([-np.pi / 2, 0, np.pi / 2]), 10)
array([[ 0.,  2.],
       [ 2.,  0.],
       [ 0., -2.]])

>>> det.surface_deriv(np.zeros((4, 5))).shape
(4, 5, 2)