nloptSolver
Return or display available NLOPT algorithms with configuration details
Syntax
enum = nloptSolver(name)
[enum,ineq,eq,uncon,glob,deriv,subprob] = nloptSolver(name)
name = nloptSolver(enum)
nloptSolver()
Description
enum = nloptSolver(name) returns the enumeration for a given NLOPT algorithm name.
[enum,ineq,eq,uncon,glob,deriv,subprob] = nloptSolver(name) also returns the following information:
ineq - 1 if the algorithm supports nonlinear
inequalities
eq - 1 if the algorithm supports nonlinear equalities
uncon - 1 if the algorithm allows unconstrained optimization (or infinite
bounds)
glob - 1 if the algorithm is a global optimization strategy
deriv - 1 if the algorithm requires a gradient and Jacobian
subprob - 1 if the algorithm requires a suboptimizer, such as the Augmented
Lagrangian method.
name = nloptSolver(enum) returns the algorithm name for a given NLOPT enumeration.
nloptSolver() prints a list of all algorithms and their configuration details.
Example
The following examples show the various uses of this function. Note the example lists below will change depending on the version you have downloaded.
Checking all solvers
To see a list of all algorithms available from NLOPT together with configuration details:
>> nloptSolver
--------------------------------------------------------------------------------------------------
NLOPT AVAILABLE SOLVERS:
NAME SCOPE
DERIV UNCON INEQ EQ DESCRIPTION
SUBOPT
AUGLAG | Global | Any | YES | YES |
YES | Augmented Langrangian - Top Layer
| Req |
GD_STOGO | Global | Req |
| | | StoGO
| |
GD_STOGO_RAND | Global | Req | |
| | StoGO, Randomized Search
| |
GN_CRS2_LM | Global | |
| | | Controlled Random Search,
Local Mutation | |
Returning the name of an NLOPT algorithm for a given enumeration
>> name = nloptSolver(1)
name =
GN_DIRECT_L
Copyright © 2011-2013 Jonathan Currie (I2C2)