ComputeMST: Euclidean Minimum Spanning Tree

In emstreeR: Tools for Fast Computing and Visualizing Euclidean Minimum Spanning Trees

Euclidean Minimum Spanning Tree

Description

Computes an Euclidean Minimum Spanning Tree (EMST) from the data. ComputeMST is a wrapper around the homonym function in the 'mlpack' library.

Usage

ComputeMST(x, verbose = TRUE, scale = FALSE)

Arguments

x	a numeric matrix or data.frame.
verbose	If TRUE, mutes the output from the C++ code.
scale	If TRUE, it will scale your data with scale before computing the the minimum spanning tree and the distances to be presented will refer to the scaled data.

Details

Before the computation, ComputeMST runs some checks and transformations (if needed) on the provided data using the data_check function. After the computation, it returns the 'cleaned' data plus 3 columns: from, to, and distance. Those columns show each pair of start and end points, and the distance between them, forming the Minimum Spanning Tree (MST).

Value

an object of class MST and data.frame.

Note

It is worth noting that the afore mentioned columns (from, to, and distance) have no relationship with their respective row in the output MST/data.frame object. The authors chose the data.frame format for the output rather than a list because it is more suitable for plotting the MST with the new 'ggplot2' Stat (stat_MST) provided with this package. The last row of the output at these three columns will always be the same: 1 1 0.0000000. This is because we always have n-1 edges for n points. Hence, this is done to 'complete' the data.frame that is returned.

Examples

## artifical data
set.seed(1984)
n <- 15
c1 <- data.frame(x = rnorm(n, -0.2, sd = 0.2), y = rnorm(n, -2, sd = 0.2))
c2 <- data.frame(x = rnorm(n, -1.1, sd = 0.15), y = rnorm(n, -2, sd = 0.3)) 
d <- rbind(c1, c2)
d <- as.data.frame(d)

## MST:
out <- ComputeMST(d)
out

emstreeR documentation built on Nov. 14, 2023, 5:10 p.m.