find.next.left.x: Find next X evaluation Point

In abn: Modelling Multivariate Data with Additive Bayesian Networks

Find next X evaluation Point

Description

Attempt to find the next x evaluation point using spline extrapolation traversing left from mode.

Usage

find.next.left.x(mat.xy, g.max, g.factor, x.delta, max.fact.delta)

find.next.right.x(mat.xy, g.max, g.factor, x.delta, max.fact.delta)

Arguments

mat.xy	matrix
g.max	integer. See Details.
g.factor	integer. See Details.
x.delta	integer. See Details.
max.fact.delta	integer. See Details.

Details

if new x point is more than a factor max.fact.delta (e.g. 0.2) from last evaluated point then stop here cat("evaluating node ",nodeid,": parameter:",paramid," at betafixed=",betafixed," with gvalue=",gvalue,"\n",sep=""); find the next x value left which differs from the max. gvalue by at least a factor of g.factor, searching in step sizes of x.delta subject to the constraint that if we move more than max.fact.delta*last.x then we evaluate here. Avoids big steps.

Value

integer

Functions

• find.next.right.x(): Attempt to find the next x evaluation point using spline extrapolation traversing right from mode.

abn documentation built on Nov. 3, 2023, 5:08 p.m.