Final Clusters

# Prepare R session
set.seed(55)

# Load essential libraries
library(vdemdata)
library(subspace)
library(tidyverse)

## ── Attaching packages ─────────────────────────────────────── tidyverse 1.3.1 ──

## ✓ ggplot2 3.3.5     ✓ purrr   0.3.4
## ✓ tibble  3.1.4     ✓ dplyr   1.0.7
## ✓ tidyr   1.1.4     ✓ stringr 1.4.0
## ✓ readr   2.0.2     ✓ forcats 0.5.1

## ── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
## x dplyr::filter() masks stats::filter()
## x dplyr::lag()    masks stats::lag()

library(ggthemes)
library(writexl)

# Load the V-Dem dataset
dem<-vdem %>% select(country_name, year, v2x_api,v2x_mpi,v2x_freexp_altinf,v2x_frassoc_thick,v2x_suffr,v2xel_frefair,v2x_elecoff, v2x_liberal, v2xcl_rol, v2x_jucon, v2xlg_legcon, v2x_partip, v2x_cspart,v2xdd_dd, v2xel_locelec, v2xel_regelec,v2xdl_delib, v2x_egal, v2xeg_eqprotec,v2xeg_eqaccess,v2xeg_eqdr)

# Omit entries that have missing values
dem <- na.omit(dem) 

# Copy initial dataframe for further reference
dem2<-dem
dem2$count<- seq.int(nrow(dem2))

# Create identificator variable based on country-year info
dem$id <-paste(dem$country_name, as.character(dem$year))


# Select mid-level indicators—21 in total, as well as identificator variables
dem<-dem %>% select(id,v2x_api,v2x_mpi,v2x_freexp_altinf,v2x_frassoc_thick,v2x_suffr,v2xel_frefair,v2x_elecoff, v2x_liberal, v2xcl_rol, v2x_jucon, v2xlg_legcon, v2x_partip, v2x_cspart,v2xdd_dd, v2xel_locelec, v2xel_regelec,v2xdl_delib, v2x_egal, v2xeg_eqprotec,v2xeg_eqaccess,v2xeg_eqdr)

# Generate index
rownames(dem) <- dem$id

# Create new dataframe to be used for identification purposes
id <-as.data.frame(dem$id)
id$count <- seq.int(nrow(id))

# Prepare initial dataframe for clustering
dem<-dem %>% select(-id)

# Run the ProClus model
p <-ProClus(dem,16,11)

# Identify clusters
clusters <- list()
for ( i in 1:16 ) {
  clusters[[i]] <- as.data.frame(p[[i]]$objects)
}

# Merge clusters with identifying variables 
countries <- list()
for ( i in 1:16 ) {
  countries[[i]] <- merge(id, clusters[[i]] , by.x=c("count"), by.y=c("p[[i]]$objects"))
  countries[[i]]$cluster <- i
}

# Create final dataframe with all country-years and their respective cluster
countries_final <- do.call(rbind, countries)

# Prepare dataframe for visualisation
countries_final = merge(countries_final, dem2, by.x=c("count"), by.y=c("count"))
countries_final <-countries_final %>% select(country_name,year,cluster)

# Identify CEE countries
romania <-countries_final %>%subset(country_name=="Romania")
poland <-countries_final%>%subset(country_name=="Poland")
hungary <-countries_final %>%subset(country_name=="Hungary")
germany <-countries_final %>%subset(country_name=="Germany")
slovakia <-countries_final %>%subset(country_name=="Slovakia")
czeschia <-countries_final %>%subset(country_name=="Czech Republic")

# Export dataset

write_xlsx(countries_final,"countries.xlsx")

ggplot(romania, aes(year, cluster))+ geom_line()+theme_stata()+
labs(x="Year", y="Cluster", col="")

ggplot(poland, aes(year, cluster))+ geom_line()+theme_stata()+
labs(x="Year", y="Cluster", col="")

ggplot(hungary, aes(year, cluster))+ geom_line()+theme_stata()+
labs(x="Year", y="Cluster", col="")

ggplot() + 
  geom_line(data=romania %>% filter(year>1945), aes(x=year ,y = cluster, color="Romania")) + 
  geom_line(data=hungary %>% filter(year>1945), aes(x=year, y = cluster,color="Hungary"))+
  geom_line(data=poland %>% filter(year>1945), aes(x=year, y = cluster,color="Poland"))+
  geom_line(data=slovakia %>% filter(year>1945), aes(x=year, y = cluster,color="Slovakia"))+
  geom_line(data=czeschia %>% filter(year>1945), aes(x=year, y = cluster,color="Czech Republic"))+
  labs(x="Year", y="Cluster", col="")+
  theme_stata()