buildLM: Linear Model Interface In SPOT: Sequential Parameter Optimization Toolbox

Description

This is a simple wrapper for the lm function, which fits linear models. The purpose of this function is to provide an interface as required by SPOT, to enable modeling and model-based optimization with linear models. The linear model is build with main effects. Optionally, the model is also subject to the AIC-based stepwise algorithm, using the `step` function from the `stats` package.

Usage

 `1` ```buildLM(x, y, control = list()) ```

Arguments

 `x` matrix of input parameters. Rows for each point, columns for each parameter. `y` one column matrix of observations to be modeled. `control` list of control parameters, currently only with parameters `useStep` and `formula`. The `useStep` boolean specifies whether the `step` function is used. The `formula` is passed to the lm function. Without a formula, a second order model will be built.

Value

an object of class `"spotLinearModel"`, with a `predict` method and a `print` method.

Examples

 ``` 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18``` ```## Test-function: braninFunction <- function (x) { (x[2] - 5.1/(4 * pi^2) * (x[1] ^2) + 5/pi * x[1] - 6)^2 + 10 * (1 - 1/(8 * pi)) * cos(x[1] ) + 10 } ## Create design points set.seed(1) x <- cbind(runif(20)*15-5,runif(20)*15) ## Compute observations at design points (for Branin function) y <- as.matrix(apply(x,1,braninFunction)) ## Create model fit <- buildLM(x,y,control = list(algTheta=optimLHD)) ## Print model parameters print(fit) ## Predict at new location predict(fit,cbind(1,2)) ## True value at location braninFunction(c(1,2)) ```

SPOT documentation built on Oct. 23, 2021, 1:06 a.m.