Eurovision 2016-2018 analysis

eurovision2019.csv

jury_vs_televoting.R

## Eurovision analysis, May 2019
## Richard D. Morey
## 2016-2018 Data available from https://data.world/datagraver/eurovision-song-contest-scores-1975-2018

library(dplyr)
library(tidyr)
library(ggplot2)
library(purrr)
library(openxlsx)
library(RCurl)

## Read in 2019 data
e2019_csv <- RCurl::getURL("https://gist.githubusercontent.com/richarddmorey/4bf962a8c80154228ff1b896a77951c0/raw/f267c99db693e5c1d27ef10518b43a40b0e9df6c/eurovision2019.csv")
e2019 <-read.csv(text = e2019_csv)
e2019$Year <- 2019

## Read in 2016-2018 data
openxlsx::read.xlsx("eurovision_song_contest_1975_2018v2.xlsx") %>% 
  mutate(Edition2 = substr(Edition, 5, 10),
         Jury.or.Televoting = factor(Jury.or.Televoting, levels = c("J","T"), labels = c("Jury", "Televoting"))) %>%
  filter(Year > 2015, Edition2 == "f" ) %>%
  group_by(Year, To.country, Jury.or.Televoting) %>%
  summarise(Points = sum(Points)) %>%
  spread(Jury.or.Televoting, Points) %>%
  ungroup() %>%
  split(.$Year) -> x

## Combine data
x[["2019"]] <- as_tibble(e2019)

x %>% map_dbl( ~ cor(.$Jury, .$Televoting, method = "k")) -> cors
x %>% map( ~  ggplot(., aes(x = Jury, y = Televoting, label=To.country)) + 
             geom_point() + 
             geom_abline(size = .5) +
             geom_text(aes(label=To.country),hjust=0, vjust=0) +
             coord_trans(x = "sqrt", y = "sqrt") +
             expand_limits(y=0, x=0) +
             ggtitle(label = paste0(.$Year[1], 
                                    ", Kendall correlation: ", 
                                    round(cors[as.character(.$Year[1])], 3)))
) -> year_plots
  
year_plots %>% map(print)