fix geometry issue for 2020 ZCTAs

walkerke · May 27, 2023 · bb8109f · bb8109f
1 parent 418dc32
commit bb8109f
Show file tree

Hide file tree

Showing 2 changed files with 26 additions and 4 deletions.
diff --git a/R/census.R b/R/census.R
@@ -152,14 +152,14 @@ get_decennial <- function(geography,
 
   if (geography == "zcta") geography <- "zip code tabulation area"
 
-  if (geography == "zip code tabulation area" && !is.null(state)) {
-    geography <- "zip code tabulation area (or part)"
-  }
-
   if (year == 2020 && sumfile == "pl" && geography == "zip code tabulation area") {
     stop("ZCTAs are not available in the 2020 PL file.", call. = FALSE)
   }
 
+  if (geography == "zip code tabulation area" && !is.null(state)) {
+    geography <- "zip code tabulation area (or part)"
+  }
+
   # if (geography == "zip code tabulation area" && is.null(state)) {
   #   stop("ZCTA data for the decennial Census is only available by state from tidycensus.",
   #        call. = FALSE)
@@ -377,6 +377,15 @@ get_decennial <- function(geography,
 
   }
 
+  # For ZCTAs, strip the state code from GEOID (issue #338 and #358)
+  # Should only happen if the GEOID is 7 characters
+  if (geography == "zip code tabulation area (or part)" && year == 2020 && unique(nchar(dat2$GEOID)) == 7) {
+    dat2 <- dat2 %>%
+      dplyr::mutate(
+        GEOID = stringr::str_sub(GEOID, start = 3L)
+      )
+  }
+
   if (!is.null(summary_var)) {
 
     sumdat <- suppressMessages(try(load_data_decennial(geography, summary_var, key, year,
@@ -387,6 +396,13 @@ get_decennial <- function(geography,
                                                          sumfile = "sf3", state, county, show_call = show_call)))
     }
 
+    if (geography == "zip code tabulation area (or part)" && year == 2020 && unique(nchar(dat2$GEOID)) == 7) {
+      dat2 <- dat2 %>%
+        dplyr::mutate(
+          GEOID = stringr::str_sub(GEOID, start = 3L)
+        )
+    }
+
     dat2 <- dat2 %>%
       inner_join(sumdat, by = "GEOID") %>%
       rename("summary_value" = !! summary_var,

diff --git a/R/helpers.R b/R/helpers.R
@@ -167,6 +167,12 @@ use_tigris <- function(geography, year, cb = TRUE, resolution = "500k",
       year <- 2020
     }
 
+    # We can't pull ZCTA shapes by state for 2020 but this is
+    # an available hierarchy in the DHC.  So set state to NULL.
+    if (year == 2020) {
+      state <- NULL
+    }
+
     z <- zctas(cb = cb, starts_with = starts_with, year = year,
                class = "sf", state = state, ...)