A data cleaning script for demonstration of a setup-input-transformation-output building block

scripts.R

# Setup/initialization
library(tidyverse)

## Wipe any downloaded files before
unlink('*.zip')
unlink('*.csv')

## Download raw data
download.file('https://github.com/hannesdatta/course-dprep/raw/master/content/docs/tutorials/data-preparation/data_without_duplicates.zip', 'data.zip')

## Unzip raw data
unzip('data.zip')

# Input 

## Load data sets into memory
streams <- read_csv('streams.csv')
songs <- read_csv('songs.csv')
country_codes <- read_delim('country_codes.csv', delim = ';')

# Transformation
streams_joined = streams %>% 
  left_join(songs, by = c('song_id'))

streams_joined %>% left_join(country_codes,
                      by=c('country'='country_code_2_letter'))

# Output
write_csv(streams_joined, 'streams_merged.csv')