RegPy: Regularization tools for inverse and ill-posed problems in Python

Added

Addtions to regpy.vecsps

• redesigned curve spaces: Different types of parameterized curves and (where applicable) also general curves can now be used interchangibly in forwarwd operators.

New StopRules

• L_curve, QuasiOpt, Lepskii, and Oracle stopping rules added

additionas to regpy.operators

• In Convolution operations in interpolation mode now not only complex, but also real UniformGridFct spaces can be handled.

additions to 'regpy.solvers'

• Added GeneralizedTikhonov solver as an interface to StopRules used for selecting the regularization parameter.
• More solvers now extract their parameters from setting; dual quardatic Tikhonov regularization works
• Allowed initialization of Settings with exact data and implementated noise generation methods of different types in Setting.

Changed

• Stooping of TikhonovCG now implemented by internal stoprule, which also do monitoring and allow general logical combinations