PGA: Principal geodesic analysis

In spherepc: Spherical Principal Curves

Description

This function performs principal geodesic analysis.

Usage

PGA(data, col1 = "blue", col2 = "red")

Arguments

data	matrix or data frame consisting of spatial locations with two columns. Each row represents longitude and latitude (denoted by degrees.
col1	color of data. The default is blue.
col2	color of the principal geodesic line. The default is red

Details

This function performs principal geodesic analysis.

Value

plot and a list consisting of

line	spatial locations (longitude and latitude by degrees) of points in the principal geodesic line.

Note

This function requires to load 'sphereplot', 'geosphere' and 'rgl' R package.

Author(s)

Jongmin Lee

References

Fletcher, P. T., Lu, C., Pizer, S. M. and Joshi, S. (2004). Principal geodesic analysis for the study of nonlinear statistics of shape. IEEE Transactions on Medical Imaging, 23, 995-1005.

See Also

LPG.

Examples

library(rgl)
library(sphereplot)
library(geosphere)
#### example 1: noisy half-great circle data
circle <- GenerateCircle(c(150, 60), radius = pi/2)
half.circle <- circle[circle[, 1] < 0, , drop = FALSE]
sigma <- 2
half.circle <- half.circle + sigma * rnorm(nrow(half.circle))
PGA(half.circle)

#### example 2: noisy S-shaped data 
#### The data consists of two parts: x ~ Uniform[0, 20], y = sqrt(20 * x - x^2) + N(0, sigma^2), 
#### x ~ Uniform[-20, 0], y = -sqrt(-20 * x - x^2) + N(0, sigma^2).
n <- 500                                           
x <- 60 * runif(n)
sigma <- 2                                         
y <- (60 * x - x^2)^(1/2) + sigma * rnorm(n)
simul.S1 <- cbind(x, y)
z <- -60 * runif(n)
w <- -(-60 * z - z^2)^(1/2)+ sigma * rnorm(n)
simul.S2 <- cbind(z, w)
simul.S <- rbind(simul.S1, simul.S2)
PGA(simul.S)

spherepc documentation built on Oct. 7, 2021, 9:14 a.m.