Optimizer

Full name Microsoft.VisualBasic.DataMining.Bonsai.Optimizer Assembly Microsoft.VisualBasic.DataMining.Bonsai Members 2

Lightweight numerical optimisers used by the Bonsai core. These mirror the scipy primitives the reference python code relies on: a 1-D Brent root-finder (scipy.optimize.brentq) and a bound-constrained L-BFGS (scipy.optimize.minimize(method="L-BFGS-B", jac=True)).

01 Syntax

Microsoft.VisualBasic.DataMining.Bonsai.Optimizer

02 Methods

Name	Overloads	Summary
BrentZero	1	Find a root of f in [a, b] assuming f(a) and f(b) have opposite signs.
Minimize	1	Minimise obj subject to box constraints bounds using a limited-memory BFGS with gradient projection.

03 Members

method BrentZero #

BrentZero(ScalarFunc, Double, Double, Object())

Find a root of f in [a, b] assuming f(a) and f(b) have opposite signs. Mirrors scipy.optimize.brentq with the same convergence criteria.

method Minimize #

Minimize(ObjWithGrad, Double(), List(Of ValueTuple(Of Double, Double)), Object())

Minimise obj subject to box constraints bounds using a limited-memory BFGS with gradient projection. Mirrors scipy L-BFGS-B on the small (3-variable) problems Bonsai needs.