Rory Quinlan
library(tidyverse)
library(dplyr)
library(factoextra)
library(cluster)
library(gridExtra)
# Load data
happiness_0 = read_csv("happiness_2018.csv", show_col_types = FALSE)
happiness_1 = happiness_0 %>%
mutate(Corruption = as.numeric(`Perceptions of corruption`))
# Remove NA values
happiness_2 = happiness_1 %>% drop_na()
dim(happiness_1)## [1] 156 10
happiness_1 = happiness_2 %>%
select(-c(`Overall rank`, `Country or region`,`Perceptions of corruption`)) # Correlogram
pairs(happiness_1[,1:4], pch = 19)
# Create A PCA
pc.happiness = happiness_1 %>%
select(-Score) %>%
prcomp(scale=TRUE)
pc.happiness## Standard deviations (1, .., p=6):
## [1] 1.7286165 1.1856234 0.7717936 0.7501125 0.5653144 0.3581444
##
## Rotation (n x k) = (6 x 6):
## PC1 PC2 PC3 PC4
## GDP per capita -0.5126579 0.2666787 -0.1164253 -0.17098649
## Social support -0.4720852 0.2376977 -0.1824560 0.27652670
## Healthy life expectancy -0.5038326 0.2424268 -0.1581973 -0.20850817
## Freedom to make life choices -0.3752315 -0.3544097 0.4418517 0.68603484
## Generosity -0.1219418 -0.6758505 -0.7238170 0.03330243
## Corruption -0.3237308 -0.4808653 0.4571483 -0.61568506
## PC5 PC6
## GDP per capita 0.23122805 -0.75485540
## Social support -0.76997346 0.13423480
## Healthy life expectancy 0.45787741 0.63970950
## Freedom to make life choices 0.25983295 -0.01432444
## Generosity 0.04285037 -0.03873052
## Corruption -0.27326438 0.03522586
# Visualize top contributors
fviz_contrib(pc.happiness,choice="var", axes=1, top=5)
# Graph Explantory power of variables
PRVar<- pc.happiness$sdev^2
PVE<- PRVar[1:9]/sum(PRVar)
PC=1:9
data=data.frame(PC, PVE)
ggplot(data=data, aes(x=PC, y=PVE))+
geom_line(color="navy")+
geom_point(aes(x=3,y=0.1),cex=5,color="orange",alpha=0.3)+
geom_point(color="red",cex=2)+
labs(title="Proportion of Variance Explained", x="Principal Component",y="pve")+
scale_x_continuous(breaks = 1:9)
# View scatterplot of greatest contributors from PCA
PC12 <- pc.happiness$x %>% as_tibble() %>% select(1:2)
pc.happiness$x %>% as_tibble() %>% select(PC1, PC2) %>%
bind_cols(happiness_1) %>%
ggplot(aes(x = `PC1`, y = `PC2`)) + geom_point()
# Find optimal clusters
happiness_4 = scale(happiness_1[,-1])
fviz_nbclust(PC12, kmeans, method = "gap_stat")
km_mod = kmeans(happiness_4, centers=3)
pam_mod = pam(happiness_4, 3)# Create variable for cluster number in df
PC12_cluster = happiness_1 %>% mutate(cluster=factor(pam_mod$cluster))# Compare kmean and pam
p1<-fviz_cluster(km_mod, data = happiness_4)
p2<- fviz_cluster(pam_mod, data = happiness_4)
grid.arrange(p1, p2, ncol=2)
# Create clusters with country name
PC12_cluster <- PC12 %>% mutate(cluster = factor(pam_mod$cluster), countryOrRegion = factor(happiness_2$`Country or region`))# View cluster diparity
p1<-ggplot(PC12_cluster,aes(x=cluster,y=PC1,color=cluster)) +
geom_boxplot()+ labs(title="GDP per Capita by Cluster")
# Boxplot
p2<-ggplot(PC12_cluster,aes(x=cluster,y=PC2 ,color=cluster)) +
geom_boxplot()+ labs(title="Healthy Life Expectancy by Cluster")
grid.arrange(p1, p2, ncol = 2)
# Filter and display the happiest countries
as.data.frame(PC12_cluster %>% filter(cluster == 1) %>% select(countryOrRegion))## countryOrRegion
## 1 Finland
## 2 Norway
## 3 Denmark
## 4 Iceland
## 5 Switzerland
## 6 Netherlands
## 7 Canada
## 8 New Zealand
## 9 Sweden
## 10 Australia
## 11 Austria
## 12 Ireland
## 13 Germany
## 14 Belgium
## 15 Luxembourg
## 16 United States
## 17 Israel
## 18 Malta
## 19 Qatar
## 20 Singapore
## 21 Uzbekistan
## 22 Hong Kong
## 23 Indonesia
## 24 Bhutan
## 25 Myanmar