app-mpt/scripts/treemap_summary.R

#!/usr/bin/env Rscript

library(tidyverse)
library(lubridate)

# Generate a summary considering current scope
summarize_scope <- function(x) {
    summarize (x, acidentes = n(),
               ds_agente_causador = names(which.max(table(ds_agente_causador))),
               ds_cnae_classe_cat = names(which.max(table(ds_cnae_classe_cat))),
               ds_emitente_cat = names(which.max(table(ds_emitente_cat))),
               idade = mean(idade_cat),
               obito = sum(cd_indica_obito == "N")/(n()),
               ds_natureza_lesao = names(which.max(table(ds_natureza_lesao))),
               ds_cbo = names(which.max(table(ds_cbo))),
               ds_parte_corpo_atingida = names(which.max(table(ds_parte_corpo_atingida))),
               sexo = sum(cd_tipo_sexo_empregado_cat == "Masculino")/(n()),
               ds_tipo_acidente = names(which.max(table(ds_tipo_acidente))),
               ds_tipo_local_acidente = names(which.max(table(ds_tipo_local_acidente))),
               turno = names(which.max(table(turno)))
               )
}

brasil <- read_csv2("../app/data/brasil.csv")
complete <- read_csv2("../app/data/completo.csv", na = c("NA", "", "Não informado"),
                      col_types = cols(
                          pais = col_character(),
                          regiao = col_character(),
                          uf = col_character(),
                          mesorregiao = col_character(),
                          microrregiao = col_character(),
                          municipio = col_character(),
                          st_acidente_feriado = col_character(),
                          ds_agente_causador = col_character(),
                          ano_cat = col_integer(),
                          ds_cnae_classe_cat = col_character(),
                          dt_acidente = col_date(),
                          st_dia_semana_acidente = col_character(),
                          ds_emitente_cat = col_character(),
                          hora_acidente = col_time(),
                          idade_cat = col_integer(),
                          cd_indica_obito = col_character(),
                          ds_natureza_lesao = col_character(),
                          ds_cbo = col_character(),
                          ds_parte_corpo_atingida = col_character(),
                          cd_tipo_sexo_empregado_cat = col_character(),
                          ds_tipo_acidente = col_character(),
                          ds_tipo_local_acidente = col_character()
                      ))

# Remove unnecessary columns
complete <- complete[-c(7, 9, 11, 12)]
complete <- drop_na(complete)

# Change absolute hour to "shifts"
complete <- complete %>% mutate(turno = case_when(
                                    hora_acidente >= hm("00:00") & hora_acidente  < hm("06:00") ~ "Madrugada",
                                    hora_acidente >= hm("06:00") & hora_acidente  < hm("12:00") ~ "Manhã",
                                    hora_acidente >= hm("12:00") & hora_acidente  < hm("18:00") ~ "Tarde",
                                    hora_acidente >= hm("18:00") & hora_acidente  <= hm("23:59") ~ "Noite",
                                    TRUE ~ NA_character_))

# Remove absolute hour
complete <- complete[-10]

# Summarize for every scope:
country <- group_by(complete, pais) %>% summarize_scope()
by_region <- group_by(complete, pais, regiao) %>% summarize_scope()
by_uf <- group_by(complete, regiao, uf) %>% summarize_scope()
by_meso <- group_by(complete, uf, mesorregiao) %>% summarize_scope()
by_micro <- group_by(complete, mesorregiao, microrregiao) %>% summarize_scope()
by_town <- group_by(complete, microrregiao, municipio) %>% summarize_scope()

# We need full information for towns
by_town <- brasil %>% inner_join(by_town, by = c("microrregiao", "municipio"))

# Write summaries
write_delim(country, "../app/data/treemap/sumario-pais.csv", delim = ";")
write_delim(by_region, "../app/data/treemap/sumario-regiao.csv", delim = ";")
write_delim(by_uf, "../app/data/treemap/sumario-uf.csv", delim = ";")
write_delim(by_meso, "../app/data/treemap/sumario-meso.csv", delim = ";")
write_delim(by_micro, "../app/data/treemap/sumario-micro.csv", delim = ";")
write_delim(by_town, "../app/data/treemap/sumario-municipio.csv", delim = ";")