Joins simple world development data with Covid-19 data for week 3 on Kaggle

join_world_dev_covid19_week3.R

#----------------------------------------------------------------
#  Add economic development data to train and test
#-------------------------------------------------------------------

covid_train <- read.csv("train.csv", stringsAsFactors = FALSE) %>%
  mutate(Province_State = ifelse(Province_State == "", "none", Province_State)) %>%
  unite(country_and_province, 
        Country_Region, Province_State,
        remove=FALSE)


covid_test <- read.csv("test.csv", stringsAsFactors = FALSE) %>%
  mutate(Province_State = ifelse(Province_State == "", "none", Province_State)) %>%
  unite(country_and_province, 
        Country_Region, Province_State,
        remove=FALSE)


#Summary of all countries in dataset
all_countries <- covid_train %>%
  group_by(country_and_province, Country_Region, Province_State) %>%
  summarize(cases_today = max(ConfirmedCases),
            deaths_today = max(Fatalities))

world_dev <- read_csv("World Development - Country.csv", 
                      col_types = cols_only(
                        CountryCode = col_character(),
                        ShortName = col_character(),
                        TableName = col_character(),
                        LongName = col_character(),
                        Alpha2Code = col_character(),
                        Region = col_character(),
                        IncomeGroup = col_character())) %>%
  filter(!is.na(Region), !is.na(IncomeGroup)) %>%
  mutate(IncomeGroup = gsub(":", "", IncomeGroup)) %>%
  mutate(IncomeGroup = gsub(" ", "_", IncomeGroup)) %>%
  mutate(Region = gsub("&", "", Region)) %>%
  mutate(Region = gsub(" ", "", Region)) %>%
  mutate(Region = gsub("-", "", Region)) %>%
  mutate(ShortName = ifelse(ShortName == "Korea", "Korea, South", ShortName)) %>%
  mutate(ShortName = ifelse(ShortName == "Côte d'Ivoire", "Cote d'Ivoire", ShortName)) %>%
  mutate(ShortName = ifelse(ShortName == "Czech Republic", "Czechia", ShortName)) %>%
  mutate(ShortName = ifelse(ShortName == "United States", "US", ShortName)) %>%
  mutate(ShortName = ifelse(ShortName == "The Bahamas", "Bahamas", ShortName)) %>%
  mutate(ShortName = ifelse(ShortName == "Myanmar", "Burma", ShortName)) %>%
  mutate(ShortName = ifelse(ShortName == "The Gambia", "Gambia", ShortName)) %>%
  mutate(ShortName = ifelse(ShortName == "Kyrgyz Republic", "Kyrgyzstan", ShortName)) %>%
  mutate(ShortName = ifelse(ShortName == "Lao PDR", "Laos", ShortName)) %>%
  mutate(ShortName = ifelse(ShortName == "Syrian Arab Republic", "Syria", ShortName)) %>%
  mutate(ShortName = ifelse(ShortName == "St. Lucia", "Saint Lucia", ShortName)) %>%
  mutate(ShortName = ifelse(ShortName == "Slovak Republic", "Slovakia", ShortName)) %>%
  mutate(ShortName = ifelse(ShortName == "St. Vincent and the Grenadines", "Saint Vincent and the Grenadines", ShortName)) %>%
  mutate(ShortName = ifelse(ShortName == "Swaziland", "Eswatini", ShortName)) %>%
  mutate(ShortName = ifelse(ShortName == "Macedonia", "North Macedonia", ShortName)) %>%
  mutate(ShortName = ifelse(ShortName == "St. Kitts and Nevis", "Saint Kitts and Nevis", ShortName)) %>%
  mutate(ShortName = ifelse(ShortName == "Congo", "Congo (Brazzaville)", ShortName)) %>%
  mutate(ShortName = ifelse(ShortName == "Dem. Rep. Congo", "Congo (Kinshasa)", ShortName))

all_countries_dev <- all_countries  %>%
  left_join(world_dev, by = c("Country_Region" = "ShortName")) %>%
  select(-TableName, -LongName, -Alpha2Code) %>%
  mutate(Region = ifelse(Country_Region == "China", "China", Region))

#Fixing Taiwan and Holy See
taiwan_ind <- which(all_countries_dev$Country_Region == "Taiwan*")
vatican_ind <- which(all_countries_dev$Country_Region == "Holy See")
diamond_ind <- which(all_countries_dev$Country_Region == "Diamond Princess")
zaandam_ind <- which(all_countries_dev$Country_Region == "MS Zaandam")

all_countries_dev$CountryCode[taiwan_ind] <- "TWN"
all_countries_dev$CountryCode[vatican_ind] <- "VAT"
all_countries_dev$CountryCode[diamond_ind] <- NA
all_countries_dev$CountryCode[zaandam_ind] <- NA

all_countries_dev$Region[taiwan_ind] <- "EastAsiaPacific"
all_countries_dev$Region[vatican_ind] <- "EuropeCentralAsia"
all_countries_dev$Region[diamond_ind] <- "Boat"



all_countries_dev_2 <- all_countries_dev %>%
  ungroup() %>%
  mutate(geo_region = Region) %>%
  select(country_and_province, CountryCode, geo_region, IncomeGroup) 


covid_train_augmented <- left_join(covid_train, all_countries_dev_2, by = "country_and_province")
covid_test_augmented <- left_join(covid_test, all_countries_dev_2, by = "country_and_province")


write.csv(covid_train_augmented, "train_augmented.csv", row.names = FALSE)
write.csv(covid_test_augmented, "test_augmented.csv", row.names = FALSE)