emmeans (version 1.6.2-1)

plot.emmGrid: Plot an emmGrid or summary_emm object

Description

Methods are provided to plot EMMs as side-by-side CIs, and optionally to display “comparison arrows” for displaying pairwise comparisons.

Usage

# S3 method for emmGrid
plot(x, y, type, CIs = TRUE, PIs = FALSE,
  comparisons = FALSE, colors = c("black", "blue", "blue", "red"),
  alpha = 0.05, adjust = "tukey", int.adjust = "none", intervals,
  frequentist, ...)
# S3 method for summary_emm
plot(x, y, horizontal = TRUE, CIs = TRUE, xlab, ylab,
  layout, scale = NULL, colors = c("black", "blue", "blue", "red"),
  intervals, plotit = TRUE, ...)

Arguments

x

Object of class emmGrid or summary_emm

y

(Required but ignored)

type

Character value specifying the type of prediction desired (matching "linear.predictor", "link", or "response"). See details under summary.emmGrid. In addition, the user may specify type = "scale", in which case a transformed scale (e.g., a log scale) is displayed based on the transformation or link function used. Additional customization of this scale is available through including arguments to ggplot2::scale_x_continuous in ... .

CIs

Logical value. If TRUE, confidence intervals are plotted for each estimate.

PIs

Logical value. If TRUE, prediction intervals are plotted for each estimate. If objecct is a Bayesian model, this requires frequentist = TRUE and sigma = (some value). Note that the PIs option is not available with summary_emm objects -- only for emmGrid objects. Also, prediction intervals are not available with engine = "lattice".

comparisons

Logical value. If TRUE, “comparison arrows” are added to the plot, in such a way that the degree to which arrows overlap reflects as much as possible the significance of the comparison of the two estimates. (A warning is issued if this can't be done.) Note that comparison arrows are not available with `summary_emm` objects.

colors

Character vector of color names to use for estimates, CIs, PIs, and comparison arrows, respectively. CIs and PIs are rendered with some transparency, and colors are recycled if the length is less than four; so all plot elements are visible even if a single color is specified.

alpha

The significance level to use in constructing comparison arrows

adjust

Character value: Multiplicity adjustment method for comparison arrows only.

int.adjust

Character value: Multiplicity adjustment method for the plotted confidence intervals only.

intervals

If specified, it is used to set CIs. This is the previous argument name for CIs and is provided for backward compatibility.

frequentist

Logical value. If there is a posterior MCMC sample and frequentist is non-missing and TRUE, a frequentist summary is used for obtaining the plot data, rather than the posterior point estimate and HPD intervals. This argument is ignored when it is not a Bayesian model.

...

Additional arguments passed to update.emmGrid, predict.emmGrid, or dotplot

horizontal

Logical value specifying whether the intervals should be plotted horizontally or vertically

xlab

Character label for horizontal axis

ylab

Character label for vertical axis

layout

Numeric value passed to dotplot when engine == "lattice".

scale

Object of class trans (in the scales package) to specify a nonlinear scale. This is used in lieu of type = "scale" when plotting a summary_emm object created with type = "response". This is ignored with other types of summaries.

plotit

Logical value. If TRUE, a graphical object is returned; if FALSE, a data.frame is returned containing all the values used to construct the plot.

Value

If plotit = TRUE, a graphical object is returned.

If plotit = FALSE, a data.frame with the table of EMMs that would be plotted. In the latter case, the estimate being plotted is named the.emmean, and any factors involved have the same names as in the object. Confidence limits are named lower.CL and upper.CL, prediction limits are named lpl and upl, and comparison-arrow limits are named lcmpl and ucmpl. There is also a variable named pri.fac which contains the factor combinations that are not among the by variables.

Details

If any by variables are in force, the plot is divided into separate panels. For "summary_emm" objects, the … arguments in plot are passed only to dotplot, whereas for "emmGrid" objects, the object is updated using … before summarizing and plotting.

In plots with comparisons = TRUE, the resulting arrows are only approximate, and in some cases may fail to accurately reflect the pairwise comparisons of the estimates -- especially when estimates having large and small standard errors are intermingled in just the wrong way. Note that the maximum and minimum estimates have arrows only in one direction, since there is no need to compare them with anything higher or lower, respectively. See the vignette("xplanations", "emmeans") for details on how these are derived.

If adjust or int.adjust are not supplied, they default to the internal adjust setting saved in pairs(x) and x respectively (see update.emmGrid).

Examples

# NOT RUN {
warp.lm <- lm(breaks ~ wool * tension, data = warpbreaks)
warp.emm <- emmeans(warp.lm, ~ tension | wool)
plot(warp.emm)
plot(warp.emm, by = NULL, comparisons = TRUE, adjust = "mvt", 
     horizontal = FALSE, colors = "darkgreen")

### Using a transformed scale
pigs.lm <- lm(log(conc + 2) ~ source * factor(percent), data = pigs)
pigs.emm <- emmeans(pigs.lm, ~ percent | source)
plot(pigs.emm, type = "scale", breaks = seq(20, 100, by = 10))

# Based on a summary. 
# To get a transformed axis, must specify 'scale'; but it does not necessarily
# have to be the same as the actual response transformation
pigs.ci <- confint(pigs.emm, type = "response")
plot(pigs.ci, scale = scales::log10_trans())
# }