dist_comp: Distances

In etree: Classification and Regression with Structured and Mixed-Type Data

Distances

Description

Compute pairwise distances starting from single objects containing the original univariate observations.

Usage

dist_comp(x, lp = 2)

Arguments

x

Object containing the original univariate observations. Currently available types and the form they need to have to be correctly recognized are the following: Logical: logical vectors; Numeric: numeric or integer vectors; Nominal: factors; Functions: objects of class "fdata"; Graphs: (lists of) objects of class "igraph"; Persistence diagrams: (lists of) objects with attributes(x)$names == "diagram". See Details to find out which distance is used in each case.

lp

Integer specifying which norm should be used to compute the distances for functional data.

Details

The distances used in each case are the following:

• Logical: Euclidean distance, implemented via dist();

• Numeric: Euclidean distance, implemented via dist();

• Nominal: Gower's distance, implemented via daisy();

• Functions: L^p-norm, implemented via metric.lp() with default options;

• Graphs: Edge Difference distance (Hammond et al., 2013), implemented via nd.edd();

• Persistence diagrams: Wasserstein distance, implemented via wasserstein() with default options;

Value

Object of class "dist" containing the pairwise distances.

References

D. K. Hammond, Y. Gur, and C. R. Johnson (2013). Graph diffusion distance: A difference measure for weighted graphs based on the graph laplacian exponential kernel. In 2013 IEEE Global Conference on Signal and Information Processing, pages 419-422.

Examples

# Number of observations
nobs <- 10

## Logical 
obj <- as.logical(rbinom(nobs, 1, 0.5))
d <- dist_comp(obj)

## Integer
obj <- rpois(nobs, 5)
d <- dist_comp(obj)

## Numeric
obj <- rnorm(nobs)
d <- dist_comp(obj)

## Factors
obj <- factor(letters[1:nobs])
d <- dist_comp(obj)

## Functional data
obj <- fda.usc::rproc2fdata(nobs, seq(0, 1, len = 100), sigma = 1)
d <- dist_comp(obj)

## Graphs
obj <- lapply(1:nobs, function(j) igraph::sample_gnp(100, 0.2))
d <- dist_comp(obj)

## Persistence diagrams
x <- lapply(rep(100, nobs), function(np) TDA::circleUnif(np))
obj <- lapply(x, TDA::ripsDiag, maxdimension = 1, maxscale = 3)
d <- dist_comp(obj)

etree documentation built on July 16, 2022, 9:05 a.m.