Nelder–Mead downhill simplex
Amoeba
Settings
SettingsT
Default
asTensorSettings s
centroid a
TypedTensor<Vector, MeasureProduct<optim-space, MeasureOne>>
contract c w s
Point
TypedTensor<'a, 'u>
evaluate f x
Objective
float<MeasureProduct<-logL, MeasureOne>> * Point
expand c r s
reflect c w s
replace a (arg2, arg2)
float<MeasureProduct<-logL, MeasureOne>>
shrink a f s
shrinkTowardsBest best x s
solve settings rng writeOut endCondition domain startPoint f
Random
LogEvent -> unit
Solution list -> int<MeasureProduct<iteration, MeasureOne>> -> OptimStopReason
(float<MeasureProduct<optim-space, MeasureOne>> * float<MeasureProduct<optim-space, MeasureOne>> * Constraint<MeasureProduct<optim-space, MeasureOne>>)[]
'a option
Solution list
swarm settings rng logger level endCondition numberOfAmoeba paramBounds startPoint f oldLevels
int
Domain
OptimisationTrace list
toFloatLogL (l, p)
TypedTensor<Scalar, MeasureProduct<-logL, MeasureOne>>
'a
float<MeasureProduct<-logL, MeasureOne>> * 'a
update a f s
valueOf (arg1, arg1)
Type something to start searching.