Sample of historical population analysis using Populate Data

population-analysis-sample.R

# Load dependencies
library(tidyverse)
library(httr)
library(ggplot2)
library(bbplot)

# Set seed
set.seed(14723)

# Set working directory
setwd(dirname(rstudioapi::getActiveDocumentContext()$path))

# Populate Data parameters
dataset <- "ds-censo-municipal-poblacion-derecho"
account <- Sys.getenv('POPULATE_DATA_ACCOUNT')
token <- Sys.getenv('POPULATE_DATA_TOKEN')
origin <- Sys.getenv('POPULATE_DATA_ORIGIN')

# Load 2 years of data
df <- NULL
years <- c(1877, 2011)

for(year in years){
  url <- paste0("https://data.populate.tools/", account , "/datasets/", dataset, ".csv?include=province,municipality&filter_by_year=", year)
  r <- GET(url, add_headers(authorization = paste0("Bearer ", token), origin = origin))
  partial_df <- content(r, "parsed")
  if(nrow(partial_df) == 0) { break }
  
  df <- rbind(df, partial_df)
}

# Plot differences between 1877 and 2011
df %>% group_by(municipality_name, date) %>% summarize(total = sum(value)) %>% 
  mutate(total0 = first(total), total1 = last(total)) %>% 
  select(municipality_name, total0, total1) %>% 
  distinct(municipality_name, total0, total1) %>% 
  ungroup() %>% 
  top_n(50, total0) %>% 
  ggplot(aes(y=reorder(municipality_name, total0), x=total0, xend=total1)) + 
  bbc_style() +
  geom_dumbbell(colour = "#dddddd",
                size = 3,
                colour_x = "#FAAB18",
                colour_xend = "#1380A1"
  ) +
  scale_x_continuous(breaks = c(0, 1000000, 2000000, 3000000),
                     labels = c(0, "1M", "2M", "3M"),
                     position="top")  +
  theme(
    axis.text.x = element_text(size = 11),
    axis.text.y = element_text(size = 10),
    panel.grid.major.y = element_blank(),
    panel.grid.major.x = element_line(color="#eaeaea")
  )