library(tidyverse)
## Warning: package 'tidyverse' was built under R version 3.3.3
## Loading tidyverse: ggplot2
## Loading tidyverse: tibble
## Loading tidyverse: tidyr
## Loading tidyverse: readr
## Loading tidyverse: purrr
## Loading tidyverse: dplyr
## Conflicts with tidy packages ----------------------------------------------
## filter(): dplyr, stats
## lag():    dplyr, stats
library(stringr)
##Add file import from github
GINI<- read_csv("C:/Users/ambra/Desktop/Data 607/W6/WB_GINI_INDEX.csv", na="..")
## Parsed with column specification:
## cols(
##   `<U+FEFF>Series Name` = col_character(),
##   `Series Code` = col_character(),
##   `Country Name` = col_character(),
##   `Country Code` = col_character(),
##   `1990 [YR1990]` = col_double(),
##   `2000 [YR2000]` = col_double(),
##   `2007 [YR2007]` = col_double(),
##   `2008 [YR2008]` = col_double(),
##   `2009 [YR2009]` = col_double(),
##   `2010 [YR2010]` = col_double(),
##   `2011 [YR2011]` = col_double(),
##   `2012 [YR2012]` = col_double(),
##   `2013 [YR2013]` = col_double(),
##   `2014 [YR2014]` = col_double(),
##   `2015 [YR2015]` = col_character(),
##   `2016 [YR2016]` = col_character()
## )
GINI<- GINI[,-c(1,2,15,16)]

##First, I will clean up the col names to show only years
colnames(GINI) <- gsub("\\s\\[YR\\d{4}\\]", "", colnames(GINI))

##Gather and select only complete cases
GINI.long<- GINI %>%  gather("Year", "Rate", 3:ncol(.), na.rm=T) %>% arrange(`Country Name`)

##As per Joel's post, "A potential analysis that could be performed is the trend in the GINI coefficient for each country (or continent) or the average of the GINI coefficient." Let's start by adding a new variable for the continent.

library(countrycode)
## Warning: package 'countrycode' was built under R version 3.3.3
continents<- countrycode(GINI.long$`Country Name`, 'country.name', 'continent')
## Warning in countrycode(GINI.long$`Country Name`, "country.name", "continent"): Some values were not matched unambiguously: Kosovo
GINI.long<-GINI.long %>% mutate(Continent=continents) %>% na.omit

##Compute the average by continent

avereg <- GINI.long%>% 
        group_by(Continent, Year) %>% 
        summarise(Rate= mean(Rate))

##Plotting the GINI  Trend per country and continent

ggplot(GINI.long, aes(Year,Rate,color = Continent)) +
  geom_line(aes(group=`Country Name`)) +
  geom_point(data=avereg, alpha = .6, size = 6) +
  theme_bw() +
  ggtitle("GINI Index Trend across countries and continents (source:WB)")+
  ylab("GINI Index")

##mapping the index
library(rworldmap)
## Warning: package 'rworldmap' was built under R version 3.3.3
## Loading required package: sp
## ### Welcome to rworldmap ###
## For a short introduction type :   vignette('rworldmap')
library(classInt)
## Warning: package 'classInt' was built under R version 3.3.3
avgGINI<-GINI %>% mutate(Average=round(rowMeans(.[, c(-1,-2)], na.rm=TRUE))) 
GINIMap <- joinCountryData2Map(avgGINI, joinCode = "ISO3",
  nameJoinColumn = "Country Code")
## 208 codes from your data successfully matched countries in the map
## 11 codes from your data failed to match with a country code in the map
## 35 codes from the map weren't represented in your data
##To customize the catMethod of the rworldmap and create default breaks - used the library classInt and reference code from https://rdrr.io/rforge/rworldmap/man/mapCountryData.html

classInt <- classIntervals(GINIMap$Average, n=8, style="jenks")
## Warning in classIntervals(GINIMap$Average, n = 8, style = "jenks"): var has
## missing values, omitted in finding classes
catMethod = classInt$brks

#creating palette
 op <- palette(c("greenyellow", "green", "green4", "rosybrown1", "orangered2","red2","red3","red4"))


mapCountryData(GINIMap, nameColumnToPlot="Average", catMethod = catMethod,   missingCountryCol = gray(.8), addLegend=TRUE, mapTitle="GINI Index across the world, average 1990-2014", colourPalette="palette", oceanCol = "LightBlue")

On average, the GINI Index seems to be higher in the Americas than in the rest of the world, meaning significant income inequality. The lower end of the range is Europe. Apparently, there was a significant drop in income disparity in Oceania between 2000 and 2008.

The map shows that countries in the southernmost region of Africa are characterized by the highest level of income inequality. Also, there are no low or medium income inequality countries in Central and South America.