CODEWITHSUNDEEP
r
MAMONI
MAMONI

Reputation: 21

R network for distance matrix

I am trying to plot network in R of a distance matrix where distances between the nodes should be proportion to the distance matrix value and node size should be proportion to the value for nodes.etworkin

Upvotes: 2

Views: 3347

Answers (2)

Sandro Valenzuela Diaz
Sandro Valenzuela Diaz

Reputation: 41

one solution I found is make a two column data frame using the distance matrix and added to the networkD3 package. check this link https://www.jessesadler.com/post/network-analysis-with-r/

> linksdf
    source target dist
1     6307  14749 1.334
2     6307  14778 1.334
3     6307   2089 1.329
4     6307   2690 1.341

> nodesdf
      nodeID group
6307    6307     n
6336    6336     l
6438    6438     h
6439    6439     o
7046    7046     u

forceNetwork(Links = linksdf, Nodes = nodesdf, Source = "source", 
             Target = "target", NodeID = "nodeID", Group = "group")

Or you can use the distance matrix directly (https://www.r-graph-gallery.com/254-use-igraph-input-format-networkd3/):

library(igraph)
library(networkD3)

# Create data
data=matrix(sample(0:1, 400, replace=TRUE, prob=c(0.95,0.05) ), nrow=20)

# Tell Igraph it is an adjency matrix... with default parameters
network=graph_from_adjacency_matrix(data)

# transform Igraph format in something readable by networkD3
network=igraph_to_networkD3(network)

# plot
simpleNetwork(network$links,
        height = 480,                     # height of frame area in pixels
        width = 480,
        linkDistance = 120,               # distance between node. Increase this value to have more space between nodes
        charge = -480,                    # numeric value indicating either the strength of the node repulsion (negative value) or attraction (positive value)
        fontSize = 27,                    # size of the node names
        linkColour = rgb(0.1,0.9,0.1,0.3),# colour of edges, MUST be a common colour for the whole graph
        nodeColour = "forestgreen",       # colour of nodes, MUST be a common colour for the whole graph
        opacity = 0.9,                    # opacity of nodes. 0=transparent. 1=no transparency
)

hope it can help.

Upvotes: 2

Jake Burkhead
Jake Burkhead

Reputation: 6535

As CoffeeRain says, next time please provide code that shows your work and gives anyone trying to answer insight into your thought process and where the real problem may lie. Are you looking for something like this?

library(maps)

data(us.cities)

#create distance matrix
d <- dist(us.cities[, c("lat", "long")])

#multidimensional scaling so we can plot and retain distance relationships
foo <- cmdscale(d, k = 2)

#everything is upside down and backwards
#plot(foo)

plot(-foo)
plot(-foo, cex = scale(us.cities$pop))

Upvotes: 2

Related Questions