response.surface.lsd: Generate the meta-model 3D response surface data

In LSDsensitivity: Sensitivity Analysis Tools for LSD Simulations

Generate the meta-model 3D response surface data

Description

This function produces a data object for the three-dimensional graphical representations of the meta-model response surfaces for a set of factors (parameters), including the confidence interval for the surfaces.

Usage

response.surface.lsd( data, model, sa, gridSz = 25, defPos = 2,
                      factor1 = 0, factor2 = 0, factor3 = 0 )

Arguments

data	an object created by a previous call to read.doe.lsd which contains all the required experimental data for the analysis.
model	an object created by a previous call to kriging.model.lsd or polynomial.model.lsd which contains the meta-model estimated hyper-parameters.
sa	an object created by a previous call to sobol.decomposition.lsd which contains the estimated total and conditional variances for all the meta-model factors.
gridSz	integer: the number of divisions in the 3D wire frame grid. The default is 25.
defPos	1, 2, 3: the position of the default/calibration configuration on the 3 plot sequence. The default is 2 (center position).
factor1	integer: the index of the first most-important factor: 0 = automatic selection (according to the Sobol index, the default); any other number = the selected factor index, according to DoE factor order.
factor2	integer: the index of the second most-important factor: 0 = automatic selection (according to the Sobol index, the default); any other number = the selected factor index, according to DoE factor order.
factor3	integer: the index of the third most-important factor: 0 = automatic selection (according to the Sobol index, the default); any other number = the selected factor index, according to DoE factor order.

Details

This function produces data for three different wire frame 3D plots. In the 3 plots, the x-y plan is defined by the 2 most-important factors (calculated or set by the user in sobol.decomposition.lsd) and the z axis represents the response variable chosen. The three different plots shows the response surface for three values of the third most-important factor: the minimum, the default/calibration and the maximum. The order the three response surfaces are shown is defined by defPos.

The automatically set most-important factors can be overridden by any factors chosen by the user by the usage of the arguments factor1, factor2 and factor3. This way, the response surfaces can be represented for a combination of any 3 factors (parameters) in the model.

Value

The function returns an object/list of class response containing three similar objects, one for each 3D plot, each of them comprised of:

calib	the predicted meta-model response values on each point of the 3D grid.
factor	the predicted values for each individual factor.
default	the predicted values for the default/calibration configuration.

Note

See the note in LSDsensitivity-package for step-by-step instructions on how to perform the complete sensitivity analysis process using LSD and R.

Author(s)

Marcelo C. Pereira [aut, cre] (<https://orcid.org/0000-0002-8069-2734>)

See Also

read.doe.lsd(), kriging.model.lsd(), polynomial.model.lsd(), sobol.decomposition.lsd()

Examples

# get the example directory name
path <- system.file( "extdata/sobol", package = "LSDsensitivity" )

# Steps to use this function:
# 1. define the variables you want to use in the analysis
# 2. load data from a LSD simulation saved results using read.doe.lsd
# 3. fit a Kriging (or polynomial) meta-model using kriging.model.lsd
# 4. identify the most influential factors applying sobol.decomposition.lsd
# 5. calculate the response surface for the selected factors using model.limits.lsd
# 6. plot the response surface

lsdVars <- c( "var1", "var2", "var3" )          # the definition of existing variables

dataSet <- read.doe.lsd( path,                  # data files folder
                         "Sim3",                # data files base name (same as .lsd file)
                         "var3",                # variable name to perform the sensitivity analysis
                         does = 2,              # number of experiments (data + external validation)
                         saveVars = lsdVars )   # LSD variables to keep in dataset

model <- kriging.model.lsd( dataSet )           # estimate best Kriging meta-model

SA <- sobol.decomposition.lsd( dataSet, model ) # find Sobol indexes

resp <- response.surface.lsd( dataSet,          # LSD experimental data set
                              model,            # estimated meta-model
                              SA )              # Sobol sensitivity analysis results

theta3d <- 310                                  # horizontal view angle
phi3d <- 30                                     # vertical view angle
grid3d <- 25

for( i in 1 : 3 ) {                             # do for each top factor
                                                # plot 3D grid charts
  zMat <- matrix( resp$calib[[ i ]]$mean, grid3d, grid3d, byrow = TRUE )
  zlim <- range( zMat, na.rm = TRUE )

  vt <- persp( resp$grid[[ 1 ]], resp$grid[[ 2 ]], zMat, col = "gray90",
               xlab = colnames( dataSet$doe )[ SA$topEffect[ 1 ] ], zlim = zlim,
               ylab = colnames( dataSet$doe )[ SA$topEffect[ 2 ] ], zlab = dataSet$saVarName,
               theta = theta3d, phi = phi3d, ticktype = "detailed" )
}

LSDsensitivity documentation built on July 4, 2022, 1:06 a.m.