Visualize Hierarchical cluster analysis.

Visualize Hierarchical cluster analysis hclust. This function compute distance using dist, and Hierarchical cluster analysis using hclust (from stats package or flashClust if installed), and render the tree with visNetwork, adding informations. Can also be called on a hclust or dist object. Needed packages : sparkline (graphics on tooltip), ggraph, igraph, flashClust

visHclust(object, ...)

# Default S3 method
visHclust(object, ...)

# S3 method for class 'data.frame'
visHclust(
  object,
  main = "",
  submain = "",
  footer = "",
  distColumns = NULL,
  distMethod = "euclidean",
  hclustMethod = "complete",
  cutree = 0,
  tooltipColumns = 1:ncol(object),
  colorEdges = "black",
  colorGroups = substr(rainbow(cutree), 1, 7),
  highlightNearest = TRUE,
  horizontal = FALSE,
  minNodeSize = 50,
  maxNodeSize = 200,
  nodesPopSize = TRUE,
  height = "600px",
  width = "100%",
  export = TRUE,
  ...
)

# S3 method for class 'dist'
visHclust(
  object,
  data = NULL,
  main = "",
  submain = "",
  footer = "",
  cutree = 0,
  hclustMethod = "complete",
  tooltipColumns = if (!is.null(data)) {     1:ncol(data) } else {     NULL },
  colorEdges = "black",
  colorGroups = substr(rainbow(cutree), 1, 7),
  highlightNearest = TRUE,
  horizontal = FALSE,
  minNodeSize = 50,
  maxNodeSize = 200,
  nodesPopSize = TRUE,
  height = "600px",
  width = "100%",
  export = TRUE,
  ...
)

# S3 method for class 'hclust'
visHclust(
  object,
  data = NULL,
  main = "",
  submain = "",
  footer = "",
  cutree = 0,
  tooltipColumns = if (!is.null(data)) {     1:ncol(data) } else {     NULL },
  colorEdges = "black",
  colorGroups = substr(rainbow(cutree), 1, 7),
  highlightNearest = TRUE,
  horizontal = FALSE,
  minNodeSize = 50,
  maxNodeSize = 200,
  nodesPopSize = TRUE,
  height = "600px",
  width = "100%",
  export = TRUE,
  ...
)

Arguments

object

hclust | dist | data.frame.

...

Don't use

main

Title. See visNetwork

submain

Subtitle. See visNetwork

footer

Footer. See visNetwork

distColumns

numeric, indice of columns used for compute distance. If NULL (default), keep all numeric and integer columns. If Not NULL, keep only numeric and integer columns

distMethod

character, the distance measure to be used for dist function. Default to 'euclidean'.

hclustMethod

character, the agglomeration method to be used for hclust function. Default to 'complete'.

cutree

numeric or integer, desired number of groups. Default to 0.

tooltipColumns

numeric, adding mini-graphics in tooltips using sparkline ? Indice of columns used in tooltip. All by default. So, we add boxplot / pie focus on sub-population vs all population using sparkline package. NULL to disable.

colorEdges

character, color of edges. Default to 'black'.

colorGroups

character, color for group in hexa ("#00FF00"). Default rainbow.

highlightNearest

boolean, highlight sub-tree on click ? Default to TRUE.

horizontal

boolean, default to FALSE

minNodeSize

numeric, in case of nodesPopSize, minimum size of a node. Default to 50. Else minNodeSize + maxNodeSize / 2.

maxNodeSize

numeric, in case of nodesPopSize, maximum size of a node. Default to 200. Else minNodeSize + maxNodeSize / 2.

nodesPopSize

boolean, nodes sizes depends on population ? Default to TRUE.

height

character, default to "600px"

width

character, default to "100%"

export

boolean, add button for export. Default to TRUE

data

data.frame, data.frame with data. Only for hclust or dist object.

Examples

if (FALSE) { # \dontrun{

#--------------
# data.frame
#--------------

# default call on data.frame
visHclust(iris, cutree = 3, colorEdges = "red")

# update some parameters
visHclust(iris, cutree = 3, tooltipColumns = c(1, 5),
  colorGroups = c("red", "blue", "green"), horizontal = TRUE)
  
# no graphics on tooltip
visHclust(iris, cutree = 3, tooltipColumns = NULL,
  main = "Hclust on iris")
  
# Title(s)
visHclust(iris, cutree = 3,  main ="My_title",
          submain = "My_sub_title", footer = "My_footer")
          
# Export
visHclust(iris, cutree = 3, export = TRUE)


# update group / individual nodes
visHclust(iris, cutree = 8) %>% 
 visGroups(groupname = "group", color ="black", 
   shape = "triangleDown", size = 75)  %>% 
 visGroups(groupname = "individual", 
   font = list(size = 150),
   color = list(background = "white", border = "purple", 
            highlight = "#e2e9e9", hover = "orange"), shape = "box") 

#--------------
# dist
#--------------

# without adding data & info in tooltip
visHclust(dist(iris[,1:4]), cutree = 3)
  
# adding data & info in tooltip
visHclust(dist(iris[,1:4]), cutree = 3, data = iris)

#--------------
# hclust
#--------------

# without adding data & info in tooltip
visHclust(hclust(dist(iris[,1:4])), cutree = 3)
  
# adding data & info in tooltip
visHclust(hclust(dist(iris[,1:4])), cutree = 3, data = iris) 
    
} # }