plot.relationshipMatrix: Heatmap for relationship Matrix

Description Usage Arguments Author(s) Examples

View source: R/plot.relationshipMatrix.r

Description

Visualization for objects of class relationshipMatrix using a heatmap of pairwise relatedness coefficients.

Usage

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
## S3 method for class 'relationshipMatrix'
plot(
  x,
  y = NULL,
  levelbreaks = NULL,
  axes = TRUE,
  cols = NULL,
  groupLines = NULL,
  ...
)

Arguments

x

Object of class relationshipMatrix

y

Optional for comparisons of objects of class relationshipMatrix

levelbreaks

list with one element for x andy. Define breaks in the color scheme of the plot. If you make to many breaks, the color scheme repeats! If y=NULL this can be an vector. If you like to have the same breaks or both relationship matrices, you can also use just one vector.

axes

a logical value indicating whether axes should be drawn on the plot. Default is TRUE.

cols

a list with one element for each relationship matrix. Colors and the number of levelbreaks should fit. But also if not, a plot is drawn. In case option y=NULL, cols can bei a vector.

groupLines

add positions to make groups more visible

...

further graphical arguments passed to function levelplot in package lattice. To create equal colorkeys for two heatmaps, use at=seq(from,to,length=9).

Author(s)

Valentin Wimmer and Hans-Juergen Auinger

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
# small pedigree
ped <- simul.pedigree(gener = 4, 7)
gp <- create.gpData(pedigree = ped)
A <- kin(gp, ret = "add")
plot(A)

# big pedigree
## Not run: 
library(synbreedData)
data(maize)
K <- kin(maize, ret = "kin")
U <- kin(codeGeno(maize), ret = "realized") / 2
# equal colorkeys
plot(K, levelbreaks = seq(0, 2, length = 9))
plot(U, levelbreaks = seq(0, 2, length = 9))

## End(Not run)

synbreed documentation built on March 12, 2021, 3:01 a.m.