pyhelios.ParameterOptimization.LBFGS Class Reference

L-BFGS gradient-based optimization. More...

Detailed Description

L-BFGS gradient-based optimization.

Requires an NLopt build that includes the LGPL Luksan solvers. PyHelios builds without them by default to keep the distributed library MIT-licensed, so this is typically unavailable – use :class:Adam or :class:BOBYQA.

Definition at line 375 of file ParameterOptimization.py.

Static Public Attributes
int	max_iterations = 200
float	ftol_rel = 1e-6
float	xtol_rel = 1e-6
bool	verify_gradients = False
float	fd_step = 1e-5

Protected Member Functions
	_to_struct (self)

Member Function Documentation

_to_struct()

pyhelios.ParameterOptimization.LBFGS._to_struct ( self )

protected

Definition at line 383 of file ParameterOptimization.py.

Member Data Documentation

fd_step

float pyhelios.ParameterOptimization.LBFGS.fd_step = 1e-5

static

Definition at line 381 of file ParameterOptimization.py.

ftol_rel

float pyhelios.ParameterOptimization.LBFGS.ftol_rel = 1e-6

static

Definition at line 378 of file ParameterOptimization.py.

max_iterations

int pyhelios.ParameterOptimization.LBFGS.max_iterations = 200

static

Definition at line 377 of file ParameterOptimization.py.

verify_gradients

bool pyhelios.ParameterOptimization.LBFGS.verify_gradients = False

static

Definition at line 380 of file ParameterOptimization.py.

xtol_rel

float pyhelios.ParameterOptimization.LBFGS.xtol_rel = 1e-6

static

Definition at line 379 of file ParameterOptimization.py.

---

The documentation for this class was generated from the following file:

• pyhelios/ParameterOptimization.py