Delimit groups in morphospace

Project convex hulls or confidence ellipses enclosing a priori groups into an existing morphospace.

proj_groups(
  mspace,
  shapes = NULL,
  groups = NULL,
  ellipse = FALSE,
  conflev = 0.95,
  density = TRUE,
  pipe = TRUE,
  ...
)

Arguments

mspace

An "mspace" object.

shapes

Optional shape data (if NULL, the function will look for them first in mspace$projected$scores and then in mspace$ordination$x).

groups

Factor; classification of observations into groups. Its length must be the same as the number of shapes provided in shapes.

ellipse

Logical; whether to plot confidence ellipses (if FALSE, convex hulls will be used instead).

conflev

Numeric; confidence level used for confidence ellipse(s).

density

Logical; whether to add density distribution for groups (univariate ordinations only).

pipe

Logical; is the function being included in a pipe?

...

Further arguments passed to hulls_by_group_2D() or ellipses_by_group_2D().

Value

If a plot device with a morphospace is open, convex hulls or confidence ellipses enclosing the scores corresponding to groups

are projected into morphospace. If pipe = TRUE the supplied "mspace" object will be modified by appending a $gr_class

slot to $projected, as well as by adding some graphical parameters (stored into the $plotinfo slot), and returned invisibly.

Details

The goal of this function is to add a classification for shapes populating the morphospace to "mspace" objects, as well as to facilitate group visualization. Other than that, it is just a wrapper for hulls_by_group_2D and ellipses_by_group_2D.

See also

hulls_by_group_2D, ellipses_by_group_2D

Examples

#load and extract relevant data, packages and information
library(magrittr)
data("shells")
shapes <- shells$shapes
species <- shells$data$species
sp_shapes <- expected_shapes(shapes, species)

#generate basic morphospace, add sampled shapes and convex hulls for species
mspace(shapes, links = links, mag = 0.7, axes = c(1,2),
       bg.model = "light gray") %>%
  proj_shapes(shapes = shapes, col = c(1:4)[species], pch = 1) %>%
  proj_groups(groups = species, col = 1:4, lwd = 2, alpha = 0.5)


#generate basic morphospace, add sampled shapes and 95% confidence for
#species
msp <- mspace(shapes, links = links, mag = 0.7, axes = c(1,2),
              bg.model = "light gray") %>%
  proj_shapes(shapes = shapes, col = c(1:4)[species], pch = 1) %>%
  proj_groups(groups = species, col = 1:4, lwd = 2, ellipse = TRUE,
              conflev = 0.95, alpha = 0.1)


#add legend using plot_mspace()
plot_mspace(msp, legend = TRUE, cex.legend = 2, pch.groups = 16)