plotd: Plot the distribution of predictions for each class

In earth: Multivariate Adaptive Regression Splines

Plot the distribution of predictions for each class

Description

Plot the distribution of the predicted values for each class. Can be used for earth models, but also for models built by lm, glm, lda, etc.

Usage

plotd(object, hist = FALSE, type = NULL, nresponse = NULL, dichot = FALSE,
      trace = FALSE, xlim = NULL, ylim = NULL, jitter = FALSE, main=NULL,
      xlab = "Predicted Value", ylab = if(hist) "Count" else "Density",
      lty = 1, col = c("gray70", 1, "lightblue", "brown", "pink", 2, 3, 4),
      fill = if(hist) col[1] else 0,
      breaks = "Sturges", labels = FALSE,
      kernel = "gaussian", adjust = 1, zero.line = FALSE,
      legend = TRUE, legend.names = NULL, legend.pos = NULL,
      cex.legend = .8, legend.bg = "white", legend.extra = FALSE,
      vline.col = 0, vline.thresh = .5, vline.lty = 1, vline.lwd = 1,
      err.thresh = vline.thresh, err.col = 0, err.border = 0, err.lwd = 1,
      xaxt = "s", yaxt = "s", xaxis.cex = 1, sd.thresh = 0.01, ...)

Arguments

To start off, look at the arguments object, hist, type.
For predict methods with multiple column responses, see the nresponse argument.
For factor responses with more than two levels, see the dichot argument.

object	Model object. Typically a model which predicts a class or a class discriminant.
hist	FALSE (default) to call density internally. TRUE to call hist internally.
type	Type parameter passed to predict. For allowed values see the predict method for your object (such as predict.earth). By default, plotd tries to automatically select a suitable value for the model in question. (This is "response" for all objects except rpart models, where "vector" is used. The choices will often be inappropriate.) Typically you would set hist=TRUE when type="class".
nresponse	Which column to use when predict returns multiple columns. This can be a column index or column name (which may be abbreviated, partial matching is used). The default is NULL, meaning use all columns of the predicted response.
dichot	Dichotimise the predicted response. This argument is ignored except for models where the observed response is a factor with more than two levels and the predicted response is a numeric vector. The default FALSE separates the response into a group for each factor. With dichot=TRUE the response is separated into just two groups: the first level of the factor versus the remaining levels.
trace	Default FALSE. Use TRUE or 1 to trace plotd — useful to see how plotd partitions the predicted response into classes. Use 2 for more details.
xlim	Limits of the x axis. The default NULL means determine these limits automatically, else specify c(xmin,xmax).
ylim	Limits of the y axis. The default NULL means determine these limits automatically, else specify c(ymin,ymax).
jitter	Jitter the histograms or densities horizontally to minimize overplotting. Default FALSE. Specify TRUE to automatically calculate the jitter, else specify a numeric jitter value.
main	Main title. Values: "string" string "" no title NULL (default) generate a title from the call.
xlab	x axis label. Default is "Predicted Value".
ylab	y axis label. Default is if(hist) "Count" else "Density".
lty	Per class line types for the plotted lines. Default is 1 (which gets recycled for all lines).
col	Per class line colors. The first few colors of the default are intended to be easily distinguishable on both color displays and monochrome printers.
fill	Fill color for the plot for the first class. For hist=FALSE, the default is 0, i.e., no fill. For hist=TRUE, the default is the first element in the col argument.
breaks	Passed to hist. Only used if hist=TRUE. Default is "Sturges". When type="class", setting breaks to a low number can be used to widen the histogram bars
labels	TRUE to draw counts on the hist plot. Only used if hist=TRUE. Default is FALSE.
kernel	Passed to density. Only used if hist=FALSE. Default is "gaussian".
adjust	Passed to density. Only used if hist=FALSE. Default is 1.
zero.line	Passed to plot.density. Only used if hist=FALSE. Default is FALSE.
legend	TRUE (default) to draw a legend, else FALSE.
legend.names	Class names in legend. The default NULL means determine these automatically.
legend.pos	Position of the legend. The default NULL means position the legend automatically, else specify c(x,y).
cex.legend	cex for legend. Default is .8.
legend.bg	bg color for legend. Default is "white".
legend.extra	Show (in the legend) the number of occurrences of each class. Default is FALSE.
vline.thresh	Horizontal position of optional vertical line. Default is 0.5. The vertical line is intended to indicate class separation. If you use this, don't forget to set vline.col.
vline.col	Color of vertical line. Default is 0, meaning no vertical line.
vline.lty	Line type of vertical line. Default is 1.
vline.lwd	Line width of vertical line. Default is 1.
err.thresh	x axis value specifying the error shading threshold. See err.col. Default is vline.thresh.
err.col	Specify up to three colors to shade the "error areas" of the density plot. The default is 0, meaning no error shading. This argument is ignored unless hist=FALSE. If there are more than two classes, err.col uses only the first two. This argument is best explained by running an example: data(etitanic) earth.mod <- earth(survived ~ ., data=etitanic) plotd(earth.mod, vline.col=1, err.col=c(2,3,4)) The three areas are (i) the error area to the left of the threshold, (ii) the error area to the right of the threshold, and, (iii) the reducible error area. If less than three values are specified, plotd re-uses values in a sensible manner. Use values of 0 to skip areas. Disjoint regions are not handled well by the current implementation.
err.border	Borders around the error shading. Default is 0, meaning no borders, else specify up to three colors.
err.lwd	Line widths of borders of the error shading. Default is 1, else specify up to three line widths.
xaxt	Default is "s". Use xaxt="n" for no x axis.
yaxt	Default is "s". Use yaxt="n" for no y axis.
xaxis.cex	Only used if hist=TRUE and type="class". Specify size of class labels drawn on the x axis. Default is 1.
sd.thresh	Minimum acceptable standard deviation for a density. Default is 0.01. Densities with a standard deviation less than sd.thresh will not be plotted (a warning will be issued and the legend will say "not plotted").
...	Extra arguments passed to the predict method for the object.

Note

This function calls predict with the data originally used to build the model, and with the type specified above. It then separates the predicted values into classes, where the class for each predicted value is determined by the class of the observed response. Finally, it calls density (or hist if hist=TRUE) for each class-specific set of values, and plots the results.

This function estimates distributions with the density and hist functions, and also calls plot.density and plot.histogram. For an overview see Venables and Ripley MASS section 5.6.

Partitioning the response into classes

Considerable effort is made to partition the predicted response into classes in a sensible way. This is not always possible for multiple column responses and the nresponse argument should be used where necessary. The partitioning details depend on the types and numbers of columns in the observed and predicted responses. These in turn depend on the model object and the type argument.

Use the trace argument to see how plotd partitions the response for your model.

Degenerate densities

A message such as
Warning: standard deviation of "male" density is 0, density is degenerate?
means that the density for that class will not be plotted (the legend will say "not plotted").

Set sd.thresh=0 to get rid of this check, but be aware that histograms (and sometimes x axis labels) for degenerate densities will be misleading.

Using plotd for various models

This function is included in the earth package but can also be used with other models.

Example with glm:

      library(earth); data(etitanic)
      glm.model <- glm(sex ~ ., data=etitanic, family=binomial)
      plotd(glm.model)
  

Example with lm:

      library(earth); data(etitanic)
      lm.model <- lm(as.numeric(sex) ~ ., data=etitanic)
      plotd(lm.model)
  

Using plotd with lda or qda

The plotd function has special handling for lda (and qda) objects. For such objects, the type argument can take one of the following values:

"response" (default) linear discriminant
"ld" same as "response"
"class" predicted classes
"posterior" posterior probabilities

Example:

    library(MASS); library(earth); data(etitanic)
    lda.model <- lda(sex ~ ., data=etitanic)
    plotd(lda.model) # linear discriminant by default
    plotd(lda.model, type="class", hist=TRUE, labels=TRUE)
    

This handling of type is handled internally by plotd and type is not passed to predict.lda (type is used merely to select fields in the list returned by predict.lda). The type names can be abbreviated down to a single character.

For objects created with lda.matrix (as opposed to lda.formula), plotd blindly assumes that the grouping argument was the second argument.

plotd does not yet support objects created with lda.data.frame.

For lda responses with more than two factor levels, use the nresponse argument to select a column in the predicted response. Thus with the default type=NULL, (which gets automatically converted by plotd to type="response"), use nresponse=1 to select just the first linear discriminant. The default nresponse=NULL selects all columns, which is typically not what you want for lda models. Example:

    library(MASS); library(earth);
    set.seed(1)      # optional, for reproducibility
    example(lda)     # creates a model called "z"
    plot(z, dimen=1) # invokes plot.lda from the MASS package
    plotd(z, nresponse=1, hist=1) # equivalent using plotd
                                 # nresponse=1 selects first linear discr.
    

The dichot=TRUE argument is also useful for lda responses with more than two factor levels.

TODO

Handle degenerate densities in a more useful way.
Add freq argument for hist.

See Also

density, plot.density
hist, plot.histogram
earth, plot.earth

Examples

if (require(earth)) {
    old.par <- par(no.readonly=TRUE);
    par(mfrow=c(2,2), mar=c(4, 3, 1.7, 0.5), mgp=c(1.6, 0.6, 0), cex = 0.8)
    data(etitanic)
    mod <- earth(survived ~ ., data=etitanic, degree=2, glm=list(family=binomial))

    plotd(mod)

    plotd(mod, hist=TRUE, legend.pos=c(.25,220))

    plotd(mod, hist=TRUE, type="class", labels=TRUE, xlab="", xaxis.cex=.8)

    par(old.par)
}

earth documentation built on Oct. 6, 2024, 1:08 a.m.