parview: Parallel coordinates view of a prediction model or function.
In DiceView: Methods for Visualization of Computer Experiments Design and Surrogate

parview.function

R Documentation

Parallel coordinates view of a prediction model or function.

Description

Renders a parallel coordinates chart showing all input dimensions and the output simultaneously. Lines are colored by the output value (viridis palette), making it easy to identify which input regions drive high or low responses.

For model-based methods (km, Kriging, WarpKriging, glm, list) a space-filling LHS prediction grid is displayed together with the observed design points (in col_points color). Kriging-based methods also add y_low / y_up axes for the predictive confidence interval.

Two rendering engines are available via the engine argument:

"parallelPlot": Interactive htmlwidget (default), requires the parallelPlot package.
"base": Static base-R plot, no extra dependency.

Usage

## S3 method for class ''function''
parview(
  fun,
  vectorized = FALSE,
  n_points = 500,
  Xlim = c(0, 1),
  col_fun = if (!is.null(col)) col else "blue",
  col = NULL,
  Xlab = NULL,
  ylab = "y",
  engine = "parallelPlot",
  ...
)

## S3 method for class 'matrix'
parview(
  X,
  y,
  col_fun = if (!is.null(col)) col else "blue",
  col_points = if (!is.null(col)) col else "red",
  col = NULL,
  Xlab = NULL,
  ylab = NULL,
  engine = "parallelPlot",
  ...
)

## S3 method for class 'Kriging'
parview(
  Kriging_model,
  n_points = 500,
  col_fun = if (!is.null(col)) col else "blue",
  col_points = if (!is.null(col)) col else "red",
  col = NULL,
  conf_level = 0.95,
  Xlab = NULL,
  ylab = NULL,
  Xlim = NULL,
  engine = "parallelPlot",
  ...
)

## S3 method for class 'WarpKriging'
parview(
  WarpKriging_model,
  n_points = 500,
  col_fun = if (!is.null(col)) col else "blue",
  col_points = if (!is.null(col)) col else "red",
  col = NULL,
  conf_level = 0.95,
  Xlab = NULL,
  ylab = NULL,
  Xlim = NULL,
  engine = "parallelPlot",
  ...
)

## S3 method for class 'km'
parview(
  km_model,
  type = "UK",
  n_points = 500,
  col_fun = if (!is.null(col)) col else "blue",
  col_points = if (!is.null(col)) col else "red",
  col = NULL,
  conf_level = 0.95,
  Xlab = NULL,
  ylab = NULL,
  Xlim = NULL,
  engine = "parallelPlot",
  ...
)

## S3 method for class 'glm'
parview(
  glm_model,
  n_points = 500,
  col_fun = if (!is.null(col)) col else "blue",
  col_points = if (!is.null(col)) col else "red",
  col = NULL,
  conf_level = 0.95,
  Xlab = NULL,
  ylab = NULL,
  Xlim = NULL,
  engine = "parallelPlot",
  ...
)

## S3 method for class 'list'
parview(
  modelFit_model,
  n_points = 500,
  col_fun = if (!is.null(col)) col else "blue",
  col_points = if (!is.null(col)) col else "red",
  col = NULL,
  Xlab = NULL,
  ylab = NULL,
  Xlim = NULL,
  engine = "parallelPlot",
  ...
)

parview(...)

Arguments

`fun`	a function or 'predict()'-like function that returns a simple numeric, or a list(mean=...,se=...).
`vectorized`	is fun vectorized?
`n_points`	number of LHS prediction points.
`Xlim`	optional input bounds matrix (2 x D); defaults to design range.
`col_fun`	base color for the prediction color scale (HSV saturation ramp from white to this color, matching the `contourview` policy).
`col`	shorthand alias for both `col_fun` and `col_points`; overridden by the individual arguments if both are supplied.
`Xlab`	optional character vector of axis labels for inputs.
`ylab`	optional string label for the output axis.
`engine`	rendering engine: `"parallelPlot"` or `"base"`.
`...`	arguments of the relevant `parview.*` method.
`X`	the matrix of input design.
`y`	the array of output values.
`col_points`	color of observed design points.
`Kriging_model`	an object of class `"Kriging"`.
`conf_level`	confidence level for uncertainty bands shown as extra axes.
`WarpKriging_model`	an object of class `"WarpKriging"`.
`km_model`	an object of class `"km"`.
`type`	the kriging type to use for model prediction.
`glm_model`	an object of class `"glm"`.
`modelFit_model`	an object returned by `DiceEval::modelFit`.

Examples

parview(branin, Xlim = rbind(c(0, 0), c(1, 1)), engine = "base")
X <- matrix(runif(15 * 2), ncol = 2)
y <- apply(X, 1, branin)
parview(X, y, engine = "base")
if (requireNamespace("rlibkriging")) { library(rlibkriging)
  X <- matrix(runif(15 * 2), ncol = 2)
  y <- apply(X, 1, branin)
  model <- Kriging(X = X, y = as.matrix(y), kernel = "matern3_2")
  parview(model, engine = "base")
}
if (requireNamespace("rlibkriging")) { library(rlibkriging)
  X <- matrix(runif(15 * 2), ncol = 2)
  y <- apply(X, 1, branin) + 5 * rnorm(15)
  model <- WarpKriging(y = y, X = X, warping = c("affine", "affine"), kernel = "matern3_2")
  parview(model, engine = "base")
}
if (requireNamespace("DiceKriging")) { library(DiceKriging)
  X <- matrix(runif(15 * 2), ncol = 2)
  y <- apply(X, 1, branin)
  model <- km(design = X, response = y, covtype = "matern3_2")
  parview(model, engine = "base")
}
x1 <- rnorm(15); x2 <- rnorm(15)
y <- x1 + x2^2 + rnorm(15)
model <- glm(y ~ x1 + I(x2^2))
parview(model, engine = "base")
if (requireNamespace("DiceEval")) { library(DiceEval)
  X <- matrix(runif(15 * 2), ncol = 2)
  y <- apply(X, 1, branin)
  model <- modelFit(X, y, type = "StepLinear")
  parview(model, engine = "base")
}
## Static base-R plot
parview(branin, Xlim = rbind(c(0, 0), c(1, 1)), engine = "base")

## Design points only
X <- matrix(runif(30 * 2), ncol = 2)
y <- apply(X, 1, branin)
parview(X, y, engine = "base")

DiceView documentation built on June 17, 2026, 5:07 p.m.