Separate data from a variable

The sub_str function of the stringr package allows us to separate a string in R according to its number of characters and their positions. For example, for Açucena - MG , we have 12 characters:

Açucena - MG
123456789012

(the dozens were omitted for obvious reasons)

To separate the city from the states, just take the string Município and remove the last 5 characters: espaço , - , espaço , estado . As the string size varies by city, I have created a function to automate this:

separarCidade <- function(x){

  n <- nchar(x)  
  cidade <- str_sub(x, 1, n-5)

  return(cidade)

}

head(separarCidade(unique(Tab1612SojaRend$Município)), 10)
 [1] "Abadia dos Dourados" "Abaeté"              "Abre Campo"         
 [4] "Acaiaca"             "Açucena"             "Água Boa"           
 [7] "Água Comprida"       "Aguanil"             "Águas Formosas"     
[10] "Águas Vermelhas"

I used head and unique above just to show how the function operates in different city names. In your case, the correct one is to separarCidade(Tab1612SojaRend$Município) to separate the names of the cities that appear repeated.

What the separarCidade function does is to calculate the number of characters in each county, and then remove the substring from the beginning to the nth-5 character.

This function separates the city and the state even if the city has '-' hyphen and has as output a list with cities and states. The test vector has a fictitious city, with hyphens where they do not actually exist.

s <- c('Abadia-dos-Dourados - MG', 'Açucena - MG')


separar <- function(x, sep = "-"){
  separador <- paste0(sep, "[^-]+$")
  cidade <- sub(separador, "", x)
  cidade <- trimws(cidade)
  separador <- paste0("^.*", sep, "([^-]+$)")
  estado <- sub(separador, '\1', x)
  estado <- trimws(estado)
  list(cidade = cidade, estado = estado)
}

separar(s)
#$cidade
#[1] "Abadia-dos-Dourados" "Açucena"            
#
#$estado
#[1] "MG" "MG"