pdhg_stepsize¶
- odl.solvers.nonsmooth.primal_dual_hybrid_gradient.pdhg_stepsize(L, tau=None, sigma=None)[source]¶
Default step sizes for
pdhg.- Parameters:
- L
Operatoror float Operator or norm of the operator that are used in the
pdhgmethod. If it is anOperator, the norm is computed withOperator.norm(estimate=True).- taupositive float, optional
Use this value for
tauinstead of computing it from the operator norms, see Notes.- sigmapositive float, optional
The
sigmastep size parameters for the dual update.
- L
- Returns:
- taufloat
The
taustep size parameter for the primal update.- sigmatuple of float
The
sigmastep size parameter for the dual update.
Notes
To guarantee convergence, the parameters
,
and
need to satisfy
This function has 4 options,
/
given or not
given.Neither
nor
are given, they are chosen as
If only
is given,
is set to
If only
is given,
is set
to
If both are given, they are returned as-is without further validation.