earth: plot.earth – R documentation

Become an expert in R — Interactive courses, Cheat Sheets, certificates and more!

plot.earth

Plot an earth object

Description

Plot an earth object. By default the plot shows model selection, cumulative distribution of the residuals, residuals versus fitted values, and the residual QQ plot.

This function calls plotres internally. The first arguments are identical to plotres.

Usage

## S3 method for class 'earth'
plot(x = stop("no 'x' argument"),

  # the following are identical to plotres arguments

  which = 1:4, info = FALSE, versus = 1, standardize = FALSE, delever = FALSE,
  level = 0, id.n = 3, labels.id = NULL, smooth.col = 2, grid.col = 0,
  jitter = 0, do.par = NULL, caption = NULL,
  trace = 0, npoints = 3000, center = TRUE, type = NULL, nresponse = NA,

  # the following are earth specific

  col.cv = "lightblue", col.grsq = 1, col.rsq = 2, col.infold.rsq = 0,
  col.mean.infold.rsq = 0, col.mean.oof.rsq = "palevioletred",
  col.npreds = if(is.null(object$cv.oof.rsq.tab)) 1 else 0, col.oof.labs = 0,
  col.oof.rsq = "mistyrose2", col.oof.vline = col.mean.oof.rsq,
  col.pch.cv.rsq = 0, col.pch.max.oof.rsq = 0, col.vline = col.grsq,
  col.vseg = 0, lty.grsq = 1, lty.npreds = 2, lty.rsq = 5, lty.vline = "12",
  legend.pos = NULL, ...)

earth_plotmodsel( # for internal use by plotres
  x, col.rsq = 2, col.grsq = 1, col.infold.rsq = 0,
  col.mean.infold.rsq = 0, col.mean.oof.rsq = "palevioletred",
  col.npreds = NULL, col.oof.labs = 0, col.oof.rsq = "mistyrose2",
  col.oof.vline = col.mean.oof.rsq, col.pch.cv.rsq = 0,
  col.pch.max.oof.rsq = 0, col.vline = col.grsq, col.vseg = 0,
  lty.grsq = 1, lty.npreds = 2, lty.rsq = 5, lty.vline = "12",
  legend.pos=NULL, add = FALSE, jitter = 0,
  max.nterms = length(object$rss.per.subset),
  max.npreds=max(1,get.nused.preds.per.subset(object$dirs,object$prune.terms)),
  ...)

Arguments

`x`	An `earth` object. This is the only required argument. (It is called "x" for consistency with the generic `plot`.)
`which,info,versus`	These arguments are identical to `plotres`. Please see the help page for `plotres`.
`standardize,delever,level`	.
`id.n,labels.id,smooth.col`	.
`grid.col,jitter`	.
`do.par,caption,trace`	.
`npoints,center`	.
`type,nresponse`	.
`col.cv`	Default `"lightblue"`. Color of cross validation line in the residuals plot. This is the residual of the mean out-fold-predicted value. The following arguments are for the model selection plot.
`col.grsq`	Default `1`. Color of GRSq line in the Model Selection plot. Use `0` for no GRSq line.
`col.rsq`	Default `2`. Color of the RSq line in the Model Selection plot. Use `0` for no RSq line.
`col.infold.rsq`	Color of in-fold RSq lines for each fold in the Model Selection plot. Applies only if `nfold` and `keepxy` were used in the original call to `earth`. Default is `0`, lines not plotted.
`col.mean.infold.rsq`	Color of mean in-fold RSq for each number of terms in the Model Selection plot. Default is `0`, line not plotted. Applies only if `nfold` and `keepxy` were used in the original call to `earth`.
`col.mean.oof.rsq`	Default `"palevioletred"`. Color of mean out-of-fold RSq for each number of terms in the Model Selection plot. Applies only if `nfold` and `keepxy` were used in the original call to `earth`. Use `0` to not plot this line.
`col.npreds`	Color of the "number of predictors" plot in the Model Selection plot. The default displays the number of predictors unless the `oof.rsq`'s are displayed. Use `0` for no "number of predictors" plot.
`col.oof.labs`	Color of fold number labels on the `oof.rsq` lines. Default is `0`, no labels.
`col.oof.rsq`	Color of out-of-fold RSq lines for each fold in the Model Selection plot. Applies only if `nfold` and `keepxy` were used in the original call to `earth`. Default is `"mistyrose2"`, a pale pink. Use `0` to not plot these lines. May be a vector of colors, which will be recycled if necessary.
`col.oof.vline`	Color of vertical line at the maximum `oof.rsq` in the Model Selection plot. Default is `col.mean.oof.rsq`.
`col.pch.cv.rsq`	Color of point plotted on the `oof.rsq` line to indicate the `cv.rsq`. for that fold (i.e., it is plotted at the number of terms selected by the in-fold GCV). Default is `0`, point not plotted.
`col.pch.max.oof.rsq`	Color of point plotted on the `oof.rsq` line to indicate the maximum `oof.rsq` for that fold. Default is `0`, point not plotted.
`col.vline`	Color of the vertical line at selected model in the Model Selection plot. Default is `col.grsq`. This will be at the maximum GRSq unless `pmethod="none"`. Use `0` for no vertical line.
`col.vseg`	Default is `0`. Color of triangular marker at top of vertical line for best GRSq.
`lty.grsq`	Line type of GRSq line in the Model Selection plot. Default is `1`
`lty.npreds`	Line type of the "number of predictors" plot in the Model Selection plot. Default is `2`.
`lty.rsq`	Line type of RSq line in the Model Selection plot. Default is `5`.
`lty.vline`	Line type of vertical line at selected model in the Model Selection plot. Default is `"12"`.
`legend.pos`	Position of the legend in the Model Selection plot. Default is `NULL` meaning automatic. Use `legend.pos=NA` or `0` for no legend. Can be something like `legend.pos="topleft"` or `legend.pos=c(6, .75)`.
`add, max.nterms, max.npreds`	`earth_plotmodsel` arguments for internal use by `plotres`.
`...`	Please see `plotres` for the details on the dots arguments. The `ylim` argument is treated specially in the model selection plot: `ymin` equal to `-1` means use the smallest GRSq or RSq value, excluding the intercept, and `ymax` equal `-1` means use the largest GRSq or RSq value.

Note

For details on interpreting the graphs, please see the earth package vignettes “Notes on the earth package” and “Variance models in earth”.

Note that cross-validation data will not be displayed unless both nfold and keepxy were used in the original call to earth.

To remove the Number of used predictors from the Model Selection graph (to reduce clutter), use col.npreds=0.

earth_plotmodsel is provided for use by plotres.

Examples

data(ozone1)
earth.mod <- earth(O3 ~ ., data = ozone1, degree = 2)
plot(earth.mod)

earth

Multivariate Adaptive Regression Splines

v5.3.0

GPL-3

Authors

Stephen Milborrow. Derived from mda:mars by Trevor Hastie and Rob Tibshirani. Uses Alan Miller's Fortran utilities with Thomas Lumley's leaps wrapper.

Initial release