R Graphical Manual

Browse All

Last data update: 2014.03.03

R: Plot a contour view of a kriging model, including design...

contourview.km

R Documentation

Plot a contour view of a kriging model, including design points

Description

Plot a contour view of a kriging model: mean response surface, fitted points and confidence surfaces. Provide a better understanding of the kriging model behaviour.

Usage

  contourview.km(model, type = "UK", center = NULL,
    axis = NULL, npoints = 20, nlevels = 10,
    col_points = "red", col_surf = "blue", filled = FALSE,
    bg_blend = 1, mfrow = NULL, Xname = NULL, yname = NULL,
    Xscale = 1, yscale = 1, xlim = NULL, ylim = NULL,
    title = NULL, add = FALSE, ...)

Arguments

`model`	an object of class `"km"`.
`type`	the kriging type to use for model prediction.
`center`	optional coordinates (as a list or data frame) of the center of the section view if the model's dimension is > 2.
`axis`	optional matrix of 2-axis combinations to plot, one by row. The value `NULL` leads to all possible combinations i.e. `choose(D, 2)`.
`npoints`	an optional number of points to discretize plot of response surface and uncertainties.
`col_points`	color of points.
`col_surf`	color for the surface.
`filled`	use filled.contour
`nlevels`	number of contour levels to display.
`bg_blend`	an optional factor of alpha (color channel) blending used to plot design points outside from this section.
`mfrow`	an optional list to force `par(mfrow = ...)` call. The default value `NULL` is automatically set for compact view.
`xlim`	an optional list to force x range for all plots. The default value `NULL` is automatically set to include all design points.
`ylim`	an optional list to force y range for all plots. The default value `NULL` is automatically set to include all design points (and their 1-99 percentiles).
`Xname`	an optional list of string to overload names for X.
`yname`	an optional string to overload name for y.
`Xscale`	an optional factor to scale X.
`yscale`	an optional factor to scale y.
`title`	an optional overload of main title.
`add`	to print graphics on an existing window.
`...`	further arguments passed to the first call of `plot3d`.

Details

Experimental points are plotted with fading colors. Points that fall in the specified section (if any) have the color specified col_points while points far away from the center have shaded versions of the same color. The amount of fading is determined using the Euclidean distance between the plotted point and center. The variables chosen with their number are to be found in the X slot of the model. Thus they are 'spatial dimensions' but not 'trend variables'.

Note

The confidence bands are computed using normal quantiles and the standard error given by predict.km.

Author(s)

Yann Richet, IRSN

Examples

## A 2D example - Branin-Hoo function. See DiceKriging package manual
## a 16-points factorial design, and the corresponding response
d <- 2; n <- 16
design.fact <- expand.grid(seq(0, 1, length = 4), seq(0, 1, length = 4))
design.fact <- data.frame(design.fact); names(design.fact)<-c("x1", "x2")
y <- branin(design.fact)

## kriging model 1 : matern5_2 covariance structure, no trend, no nugget effect

m1 <- km(design = design.fact, response = y)

## the same as contourview.km
contourview(m1)

## change colors
contourview(m1, col_points = "firebrick", col_surf = "SpringGreen2")

## change colors,  use finer grid and add needles
contourview(m1, npoints = c(50, 30), col_points = "orange",
col_surf = "SpringGreen2")

## Display reference function
contourview(branin,dim=2,add=TRUE,col='red')

Results


R version 3.3.1 (2016-06-21) -- "Bug in Your Hair"
Copyright (C) 2016 The R Foundation for Statistical Computing
Platform: x86_64-pc-linux-gnu (64-bit)

R is free software and comes with ABSOLUTELY NO WARRANTY.
You are welcome to redistribute it under certain conditions.
Type 'license()' or 'licence()' for distribution details.

R is a collaborative project with many contributors.
Type 'contributors()' for more information and
'citation()' on how to cite R or R packages in publications.

Type 'demo()' for some demos, 'help()' for on-line help, or
'help.start()' for an HTML browser interface to help.
Type 'q()' to quit R.

> library(DiceView)
Loading required package: DiceKriging
Loading required package: DiceEval
Loading required package: rgl
> png(filename="/home/ddbj/snapshot/RGM3/R_CC/result/DiceView/contourview.km.Rd_%03d_medium.png", width=480, height=480)
> ### Name: contourview.km
> ### Title: Plot a contour view of a kriging model, including design points
> ### Aliases: contourview.km
> ### Keywords: models
> 
> ### ** Examples
> 
> ## A 2D example - Branin-Hoo function. See DiceKriging package manual
> ## a 16-points factorial design, and the corresponding response
> d <- 2; n <- 16
> design.fact <- expand.grid(seq(0, 1, length = 4), seq(0, 1, length = 4))
> design.fact <- data.frame(design.fact); names(design.fact)<-c("x1", "x2")
> y <- branin(design.fact)
> 
> ## kriging model 1 : matern5_2 covariance structure, no trend, no nugget effect
> 
> m1 <- km(design = design.fact, response = y)

optimisation start
------------------
* estimation method   : MLE 
* optimisation method : BFGS 
* analytical gradient : used
* trend model : ~1
* covariance model : 
  - type :  matern5_2 
  - nugget : NO
  - parameters lower bounds :  1e-10 1e-10 
  - parameters upper bounds :  2 2 
  - best initial criterion value(s) :  -82.14532 

N = 2, M = 5 machine precision = 2.22045e-16
At X0, 0 variables are exactly at the bounds
At iterate     0  f=       82.145  |proj g|=       1.0845
At iterate     1  f =       81.844  |proj g|=       0.99914
At iterate     2  f =       81.375  |proj g|=        1.3577
At iterate     3  f =       81.062  |proj g|=       0.35041
At iterate     4  f =       81.058  |proj g|=       0.10154
At iterate     5  f =       81.058  |proj g|=      0.004384
At iterate     6  f =       81.058  |proj g|=    5.1665e-05
At iterate     7  f =       81.058  |proj g|=    2.5808e-08

iterations 7
function evaluations 9
segments explored during Cauchy searches 9
BFGS updates skipped 0
active bounds at final generalized Cauchy point 1
norm of the final projected gradient 2.58084e-08
final function value 81.0576

F = 81.0576
final  value 81.057643 
converged
> 
> ## the same as contourview.km
> contourview(m1)
> 
> ## change colors
> contourview(m1, col_points = "firebrick", col_surf = "SpringGreen2")
> 
> ## change colors,  use finer grid and add needles
> contourview(m1, npoints = c(50, 30), col_points = "orange",
+ col_surf = "SpringGreen2")
> 
> ## Display reference function
> contourview(branin,dim=2,add=TRUE,col='red')
> 
> 
> 
> 
> 
> dev.off()
null device 
          1 
>