dendextend (version 1.4.0)

highlight_distinct_edges: Highlight distint edges in a tree (compared to another one)

Description

Highlight distint edges in a tree (compared to another one) by changing the branches' color, line width, or line type.

This function enables this feature in dend_diff and tanglegram

Usage

highlight_distinct_edges(dend, ...) 
"highlight_distinct_edges"(dend, dend2,  value = 2,  edgePar = c("col", "lty", "lwd"),  ...)  
"highlight_distinct_edges"(dend, ..., which)

Arguments

dend

a dendrogram or dendlist to find unique edges in (to highlight)

...

Ignored.

dend2

a dendrogram to compare with

value

a new value scalar for the edgePar attribute.

edgePar

a character indicating the value inside edgePar to adjust. Can be either "col", "lty", or "lwd".

which

an integer vector indicating, in the case "dend" is a dendlist, on which of the trees should the modification be performed. If missing - the change will be performed on all of objects in the dendlist.

Value

A dendrogram with modified edges - the distinct ones are changed (color, line width, or line type)

See Also

distinct_edges, highlight_distinct_edges, dist.dendlist, tanglegram assign_values_to_branches_edgePar, distinct.edges,

Examples

x <- 1:5 %>% dist %>% hclust %>% as.dendrogram
y <- set(x, "labels", 5:1)
distinct_edges(x, y)
distinct_edges(y, x)

par(mfrow = c(1,2))
plot(highlight_distinct_edges(x, y))
plot(y)

# tanglegram(highlight_distinct_edges(x, y),y)
# dend_diff(x, y)

## Not run: 
# 
# # using  highlight_distinct_edges combined with dendlist and set
# # to clearly highlight "stable" branches.
# data(iris); ss <- c(1:5, 51:55, 101:105)
# iris1 <-iris[ss,-5] %>% dist %>% hclust(method = "single") %>% as.dendrogram 
# iris2 <- iris[ss,-5] %>% dist %>% hclust(method = "complete") %>% as.dendrogram
# iris12 <- dendlist(iris1, iris2) %>% set("branches_k_color",k=3) %>% 
#    set("branches_lwd", 3) %>%  highlight_distinct_edges(value = 1, edgePar = "lwd")
# iris12 %>% untangle(method = "step2side")  %>% 
# tanglegram(sub="Iris dataset", main_left = "'single' clustering", 
# 				main_right = "'complete' clustering")
# ## End(Not run)