In joshoriv/rivera.package: Analyze Siren Shape Data

knitr::opts_chunk$set(echo = TRUE)

Howdy! The first thing one should do to utilize this set of functions is install the Rivera package. code seen below

devtools::install_github("joshoriv/rivera.package")

Once the package has been installed, please be sure to load the package into your R console.

library(rivera.package)

At this point, we can go ahead and load the required packages for this tutorial. Following this we'll be reading in the siren locality data and siren shape data (landmark coordinates).

library(readr)
library(ggplot2)
library(sf)
library(rnaturalearth)
library(rnaturalearthdata)
library(mapdata)
library(maps)
library(rgl)
library(geomorph)
library(svd)
library(shapes)
sirens <- readr::read_csv("../inst/extdata/sirens_updated.csv")
file_path <- system.file("extdata", "SirenDorsal.TPS", package = "rivera.package")
sirens_dorsal <- geomorph::readland.tps(file_path, specID = "ID")

Specimen Locality Distribution Map

The purpose of this function is to generate a map of siren specimen localities. Specimen distribution maps are helpful in understanding trends in biological phenomena. In order for this function to process, the coordinate information (longitude & latitude) must be available for each specimen. The world map, where the specimen datapoints will be superimposed, will be selected by providing coordinates of interest.

sirens_map <- dplyr::select(sirens, Longitude, Latitude) %>% 
  na.omit()
world <- ne_countries(scale = "medium", returnclass = "sf")
ggplot(data = world) +
  geom_sf() +
  geom_point(data = sirens_map, mapping = aes(x = Longitude, y = Latitude), color = "blue") +
  coord_sf(xlim = c(-110, -75), ylim = c(22, 50), expand = FALSE)

Generalized Procrustes Analysis

This function performs a generalized Procrustes analysis/superisposition of multiple specimens about their mean. This function requires the input of specimen landmark coordinates. A plot of superimposed coordinates is returned and their centroid sizes in dataframe formate. This function then plots the Procrustres coordinates. This is a useful analysis for visualizing subtle variation in morphology.

gpa <- gpagen(sirens_dorsal)
plot(gpa)

Principal Components Analysis

This function performs a traditional principal components analysis. The input for this function is the generalized Procrustes analysis proceeding this function. A PCA plot shows clusters of samples based on their similarity. This can be helpful to visualize similarities among geographically isolated taxa, different species, etc.

pca <- gm.prcomp(gpa$coords)
  summary(pca)
  plot(pca, main = "PCA")

joshoriv/rivera.package documentation built on Jan. 1, 2021, 5:40 a.m.