Extracts Armenia 2011 census data for all settlements from pdf file issued by Armenia's statistical office

extract_armenia_census_2011.R

library("tabulizer")
library("tidyverse")

dir.create("data", showWarnings = FALSE)
dir.create(file.path("data", "original_files"), showWarnings = FALSE)

census_2011_pdf_url <- "https://www.armstat.am/file/article/1._bajin_1_182-311.pdf"
census_2011_pdf_file <- file.path("data", "original_files", "census_2011.pdf")

if (file.exists(census_2011_pdf_file)==FALSE) {
  download.file(url = census_2011_pdf_url, destfile = census_2011_pdf_file)
}

tables <- extract_tables(file = census_2011_pdf_file, pages = 36:101, output = "data.frame")

tables_cleaned <- vector("list", length(36:101)) 

for (i in seq_along(tables)) {
  if (ncol(tables[[i]])==8) {tables[[i]] <- tables[[i]][,-5]} #deals with problematic page
  tables_cleaned[[i]] <- set_names(as.data.frame(x = tables[[i]],stringsAsFactors = FALSE), nm = c("Village", "Total_de_facto", "Men_de_facto", "Women_de_facto", "Total_de_jure", "Men_de_jure", "Women_de_jure")) %>% 
  mutate(Village = lag(Village)) %>% 
  filter(Total_de_facto != "") %>% 
  mutate_all(.funs = str_remove_all, pattern = ",") %>% 
  mutate_if(str_detect(string = colnames(.), pattern = "_"), .funs = as.numeric)
}

census_2011 <- purrr::map_df(.x = tables_cleaned, .f = bind_rows)

write_csv(x = census_2011, path = file.path("data", "armenia_census_2011.csv"))

Warning: it's not 100% perfect, does not recognise three lines in the last page of table 1.3 (page 101), and some other data may be missing.