#S12 - Recap

#S12 - Recap script

This contains the script developed during the recap session.

recap.R

here::i_am("R/recap.R")
library(dplyr)
library(data.table)
library(here)
library(tidyr)
library(ggplot2)

# How to deal with large files - some hints-----------
gdp_file <- here("data/raw/API_NY.GDP.PCAP.PP.KD_DS2_en_csv_v2_5359165.csv")
gdp_data_raw <- data.table::fread(file = gdp_file)

# Use overview functions:
head(gdp_data, n = 2)
names(gdp_data)
str(gdp_data)
dplyr::glimpse(gdp_data)

# Check the unique values of the columns:
unique(gdp_data$`Indicator Name`)
unique(gdp_data$V67)

# Taking into account all this information suggests to augment the import call:
gdp_data_raw <- data.table::fread(
  file = gdp_file, 
  header = TRUE # To ensure the column names are correct
  ) %>% 
  tibble::as_tibble(.) %>% # Facilitates printing
  select( # Remove redundant columns
    -c("Country Name", 
       "Indicator Name", "Indicator Code",
       "V67")
  ) 

# Then continue working with the data:
gdp_data_tidy <- gdp_data_raw %>% 
  tidyr::pivot_longer(
    cols = -"Country Code", 
    names_to = "year", 
    values_to = "GDP_percapita")

# Country codes--------------
library(countrycode)
gdp_data_countrynames <- gdp_data_tidy %>% 
  dplyr::mutate(
    countryname = countrycode::countrycode(
    `Country Code`, origin = "iso3c", destination = "country.name")
    ) %>% 
  dplyr::mutate(# For manual correction do, e.g.:
    countryname = ifelse(`Country Code` == "WLD", "World", countryname)
  )
head(gdp_data_countrynames)
# Check potential duplicates!

# Scatter plot---------------
wine_data <- DataScienceExercises::wine2dine

ggplot(
  data = wine_data, 
  mapping = aes(
    y = `residual sugar`, 
    x = alcohol, 
    color = kind)
  ) + 
  geom_point() +
  theme_bw()