data_prep census_tract_2022

data_prep/R/census_tract_2022.R

@@ -0,0 +1,90 @@
+library(sf)
+library(data.table)
+library(dplyr)
+
+dest_dir <- 'C:/Users/user/Downloads/BR_Malha_Preliminar_2022/'
+
+df <- sf::st_read('C:/Users/user/Downloads/BR_Malha_Preliminar_2022/BR_Malha_Preliminar_2022.gpkg')
+saveRDS(df, 'C:/Users/user/Downloads/BR_Malha_Preliminar_2022/BR_Malha_Preliminar_2022.rds')
+
+df$AREA_KM2 <- NULL
+
+temp_sf <- dplyr::rename(df,
+                     code_tract = CD_SETOR,
+                     code_muni = CD_MUN,
+                     name_muni = NM_MUN,
+                     code_subdistrict = CD_SUBDIST,
+                     name_subdistrict = NM_SUBDIST,
+                     code_district = CD_DIST,
+                     name_district = NM_DIST,
+                     code_urban_concentration = CD_CONCURB,
+                     name_urban_concentration = NM_CONCURB,
+                     code_state = CD_UF,
+                     name_state = NM_UF,
+                     code_micro = CD_MICRO,
+                     name_micro = NM_MICRO,
+                     code_meso = CD_MESO,
+                     name_meso = NM_MESO,
+                     code_immediate = CD_RGI,
+                     name_immediate = NM_RGI,
+                     code_intermediate = CD_RGINT,
+                     name_intermediate = NM_RGINT,
+                     code_region = CD_REGIAO,
+                     name_region = NM_REGIAO
+                     )
+head(temp_sf)
+
+
+# remove P from code tract
+temp_sf <- mutate(temp_sf, code_tract = gsub("P","", code_tract))
+head(temp_sf)
+
+
+# make all columns as character
+char_cols <- names(temp_sf)
+char_cols <- char_cols[char_cols %like% 'code_|name_']
+temp_sf <- mutate(temp_sf, across(all_of(char_cols), as.character))
+
+
+# Use UTF-8 encoding
+temp_sf <- use_encoding_utf8(temp_sf)
+
+# Harmonize spatial projection CRS, using SIRGAS 2000 epsg (SRID): 4674
+temp_sf <- harmonize_projection(temp_sf)
+
+save_state <- function(code_uf){ # code_uf <- 11
+
+  temp_sf2 <- subset(temp_sf, code_state == code_uf)
+
+ # temp_sf2 <- subset(temp_sf2, code_muni == '3304557')
+
+
+
+
+  # convert to MULTIPOLYGON
+  temp_sf2 <- to_multipolygon(temp_sf2)
+
+  # simplify
+  temp_sf_simplified <- simplify_temp_sf(temp_sf2)
+
+  # Save cleaned sf in the cleaned directory
+  sf::st_write(temp_sf2, paste0(dest_dir, code_uf,".gpkg") )
+  sf::st_write(temp_sf_simplified, paste0(dest_dir, code_uf,"_simplified.gpkg"),)
+
+}
+
+
+all_states <- unique(temp_sf$code_state)
+
+# Apply function to save the data
+gc(reset = T)
+
+tictoc::tic()
+future::plan(strategy = 'multisession')
+furrr::future_map(.x=all_states, .f=save_state, .progress = T)
+tictoc::toc()
+
+
+tictoc::tic()
+pbapply::pblapply(X=all_states, FUN=save_state)
+tictoc::toc()

data_prep/R/support_fun.R

@@ -278,6 +278,7 @@ to_multipolygon <- function(temp_sf){
       temp_sf <- subset(temp_sf, sf::st_geometry_type(temp_sf) |> as.character() != "LINESTRING")
 
       # get polyons
+      temp_sf <- sf::st_cast(temp_sf, "POLYGON")
       temp_sf <- sf::st_collection_extract(temp_sf, "POLYGON")
       temp_sf <- sf::st_cast(temp_sf, "MULTIPOLYGON")
       return(temp_sf)