improve headwater handling

R/4_find_match.R

@@ -100,13 +100,26 @@ get_hr_pairs <- function(nhdhr_path, nhdplus_hw_outlets, prj, cores = 5, max_str
   dplyr::bind_rows(hr_pairs)
 }
 
+#' add 24k headwater pairs
+#' @param hr_pairs id COMID pairs from catchment-based headwater matching
+#' @param extra_file character path to file containing previously matched headwaters
+#' 
 add_24k_pairs <- function(hr_pairs, extra_file) {
   
   found <- readr::read_csv(extra_file, col_types = "ccncc")
   
-  found <- found[!found$hr_id %in% hr_pairs$id,]
+  # where a mainstem is in the 24k headwaters but not hr
+  found_new <- found[!found$COMID %in% hr_pairs$COMID,]
 
-  bind_rows(select(found, COMID, id = hr_id),
+  # where the 24k headwaters are different
+  found_diff <- inner_join(select(found, COMID, hr_id), hr_pairs, by = "COMID") |>
+    filter(hr_id != id)
+  
+  # remove where things are different
+  hr_pairs <- filter(hr_pairs, !COMID %in% found_diff$COMID)
+  
+  bind_rows(select(found_new, COMID, id = hr_id),
+            select(found_diff, COMID, id = hr_id),
             select(hr_pairs, COMID, id))
     
 }