Assignment 7 - 5): One-way analysis of variance

[SebastianVeuskens]

Task

A one-way analysis of variance (also called ANOVA) should be performed in this exercise.

What is the ANOVA?

A great quick guide on how to perform an ANOVA can be found here.

Code

library(fda)
library(dplyr)

# Load data 
covid21 <- read.table("data/covid2021.dat") 
covid21v <- as.matrix(covid21[,5:559]) 

# B-spline basis 
bbasis100 <- create.bspline.basis(c(1,555), nbasis=100) 

# Functional Data object
fdcovid100 <- Data2fd(1:555, y=t(as.matrix(covid21v)), basisobj=bbasis100)

# Principal Component Analysis 
covidpca1 <- pca.fd(fdcovid100, nharm = 1)
ncontinent <- as.numeric(as.factor(covid21$continent)) 

con_scores_all <- data.frame(scores = covidpca1$scores, continent = as.factor(ncontinent))

# Anova test 
res.aov <- aov(scores ~ continent, data = con_scores_all)

In this case, the null-hypothesis (all countries have the same mean score) should be rejected.