In wardfont/edcpR: Ecological Data Collection and Processing Package

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>"
)

library(edcpR)

Data

The first three exercises use the irisdataset:

library(datasets)
data(iris, pacakges = "datasets")

The last exercise uses the varespec dataset:

library(vegan)
data(varespec, pacakges = "vegan")

Hierarchical clustering

• Construct a dissimilarity matrix for the sepal length
• Perform an agglomerative hierarchical clustering of the dataset
• based upon the sepal length
• TIP: cluster package
• Plot the dendrogram of the performed clustering
• change the output from the agglomerative clustering into a hclust object
• Find the optimum number of clusters
• TIP: stats package
• How does the clustering changes when you use 4 clusters instead of 3?
• Visualize the different clusters by plotting rectangles on the dendrogram

K means

• Find the optimum number of clusters
• TIP: cluster package
• Perform a K means clustering of the iris dataset with selected number of clusters
• based upon the petal length & width
• Visualize the clustering using ggplot2

K nearest neighbours

• Split data into random training and random testing sets
• use a random 70%-30% split
• Find the optimum number of neighbours
• TIP: class package
• Visualize the clustering using ggplot2

Indicator species analysis

• Use the varespec dataset
• Divide the species data in clusters
• Find indicator species for each cluster
• TIP: indicspecies package

Assignment 5

Access the assignment:

vignette("assignment-5", package = "edcpR")

Submit your R-script as Surname_FirstName_assignment5.R

Upload everything on Toledo before December 8, 12 pm (= at noon)!

wardfont/edcpR documentation built on Dec. 23, 2021, 5:07 p.m.