etomo.operators.linear.base#

Base class for linear operators

class LinearBase(n_jobs=4, backend='threading', n_channels=1, verbose=False)[source]#

Bases: modopt.opt.linear.base.LinearParent

Creates a base class for all linear operators. Ensures that the operator is defined in a way that is compatible with modopt reconstruction algorithms and handles multithreading for multichannel reconstructions. Any linear operator in pysap-etomo inherits from this class. To create an operator based on this class, the direct and adjoint operators have to be implemented in the _op and _adj_op methods respectively.

_op(data)[source]#

_adj_op(coeffs)[source]#

op(data)[source]#

Direct operator

Parameters: data (numpy.ndarray) – 2D or 3D images to compute operator on.
Returns: coeffs – Linear coefficients of imput images
Return type: numpy.ndarray

adj_op(coeffs)[source]#

Adjoint operator

Parameters: coeffs (numpy.ndarray) – Linear coefficients
Returns: images – Reconstructed data from imput coefficients
Return type: numpy.ndarray

l2norm(shape)[source]#

Compute the L2 norm. Can be overwritten if an analytical formula exists.

Parameters: shape (tuple(int)) – the 2D or 3D data shape.
Returns: norm – the L2 norm.
Return type: float