bug in sort_network when duplicated rows due to attributes fixes #52

DESCRIPTION

@@ -1,6 +1,6 @@
 Package: hydroloom
 Title: Utilities to Weave Hydrologic Fabrics
-Version: 1.1.0
+Version: 1.1.1
 Authors@R: 
     person(given = "David", 
            family = "Blodgett", 
@@ -14,7 +14,7 @@ Suggests: testthat, nhdplusTools, future, lwgeom, future.apply, knitr, gifski, m
 License: CC0
 Encoding: UTF-8
 Roxygen: list(markdown = TRUE)
-RoxygenNote: 7.3.1
+RoxygenNote: 7.3.2
 Config/testthat/parallel: true
 Config/testthat/edition: 3
 VignetteBuilder: knitr

NEWS.md

@@ -1,3 +1,8 @@
+hydroloom 1.1.1
+==========
+
+- fix bug with sort_network when duplicate entries are in the extended attributes -- #52
+
 hydroloom 1.1.0
 ==========
 

R/make_index_ids.R

@@ -53,6 +53,11 @@ make_index_ids.hy <- function(x, long_form = FALSE) {
           Run check_hy_graph to identify issues.")
   }
 
+  vars <- c("id", "toid")
+  if("downmain" %in% names(x)) vars <- c(vars, "downmain")
+
+  x <- select(x, all_of(vars))
+
   x <- distinct(x)
 
   out_val <- get_outlet_value(x)
@@ -66,9 +71,6 @@ make_index_ids.hy <- function(x, long_form = FALSE) {
     out_rename <- copy(out)
     setnames(out_rename, old = "indid", new = "toindid")
 
-    vars <- c("id", "toid")
-    if("downmain" %in% names(x)) vars <- c(vars, "downmain")
-
     out <- merge(merge(as.data.table(x)[, vars, with = FALSE],
                        out, by = "id", all.x = TRUE, sort = FALSE),
                  out_rename,

R/sort_network.R

@@ -64,6 +64,7 @@ sort_network.data.frame <- function(x, split = FALSE, outlets = NULL) {
 #' @export
 #'
 sort_network.hy <- function(x, split = FALSE, outlets = NULL) {
+
   hy_g <- get_hyg(x, add = TRUE, id = id)
 
   x <- select(st_drop_geometry(x), id, toid, everything())

tests/testthat/test_sort_network.R

@@ -217,3 +217,17 @@ test_that("add_topo_sort deals with diversions", {
       unique(base_network$topo_sort[base_network$id == 8317403]))
 
 })
+
+test_that("duplicated attributes", {
+  network <- readRDS("data/sort_network_dups.rds")
+
+  dedup <- dplyr::distinct(dplyr::select(network, id, toid))
+
+  sorted <- sort_network(network)
+
+  sorted_dedup <- sort_network(dedup)
+
+  sorted_dedup_2 <- dplyr::distinct(dplyr::select(sorted, id, toid))
+
+  expect_equal(sorted_dedup, sorted_dedup_2)
+})