tpos_s4ag5.md

Finalising the Data Matrix

This step could have been added to previous code, but to highlight the clarity needed separate treatment. This code does nothing great except picking up each individual CSV files and transposing the rows into columns so that countries are represented as variables. The rows give discrete data on commodities.

This essentially involves spreading the data, which was in a gathered form in the previous step. And, then re-writing the files.

First step is to initialize the libraries.

library(dplyr)
library(tidyverse)

aa <- list.files(pattern = "*.csv", recursive = T)
length(aa)

for (i in aa){
  s1 <- read.csv(i, stringsAsFactors = F, strip.white = T)
  
  s2 <- s1 %>%
    select("Partner", "Code", "Commodity", "TradeValue") %>%
    spread(Partner, TradeValue)
  sapply(s2, class)
  s2[is.na(s2)] <- 0
  s2$Total <- rowSums(s2[3:ncol(s2)])
  s3 <- s2 %>% 
    group_by(Code) %>%
    top_n(1, abs(Total))
  
  s3$Total <- NULL
  
  colnames(s3)[1] <- "Product"
  write.csv(s3, i, row.names = F)
  
}

The final sets of data are ready can be examined here.