Predict: Predict from a 'GaSPModel' object.

In GaSP: Train and Apply a Gaussian Stochastic Process Model

Predict from a GaSPModel object.

Description

Predict from a GaSPModel object.

Usage

Predict(GaSP_model, x_pred, generate_coefficients = c(FALSE, TRUE))

Arguments

GaSP_model	Object of class GaSPModel.
x_pred	A data frame containing the values of the input variables at which to predict the output.
generate_coefficients	A boolean indicating whether coefficients for further external predictions are generated.

Value

A list with the following elements:

y_pred	A data frame with two columns: the predictions Pred and their standard errors SE.
pred_coeffs	A vector of coefficients for further predictions; NULL if generate_coefficients is FALSE.

Note

The vector of prediction coefficients in pred_coeffs can be used as follows. Let c denote the coefficients and let r denote a vector with element i containing the correlation between the output at a given new point and the output at training point i. Then the prediction for the output at the new point is the dot product of c and r.

RMSE computes the root mean squared error of the predictions. PlotPredictions and PlotResiduals plot the predictions or their residuals; PlotStdResiduals and PlotQQ plot the standardized residuals.

Examples

borehole_pred <- Predict(
  GaSP_model = borehole_fit,
  x_pred = borehole$x_pred,
  generate_coefficients = TRUE
)

GaSP documentation built on May 31, 2023, 7:54 p.m.