fix bug in habitat codes for resident distributions (PR #56)

- Fix bug in `create_spp_info_data()` in assigning habitat types for resident
  distributions of migratory and non-migratory species. This bug meant that (i)
  habitat types for resident distributions of migratory species would include
  those exclusively affiliated with the species' passage distributions and (ii)
  habitat types for resident distributions of non-migratory species would not
  include those exclusively affiliated with the species' passage distributions.
  Thanks to Jianqiao Zhao for bug report.
- Update built-in testing data to include multiple migratory species.
- Update internal R script for creating test dataset
  (i.e., `inst/scripts/test-data.R`) to be compatible with current version of
  the package.
- Update `is_gdal_calc_available()` to be more robust.
- Fix URLs.

DESCRIPTION

@@ -1,6 +1,6 @@
 Package: aoh
 Type: Package
-Version: 0.0.2.11
+Version: 0.0.2.12
 Title: Create Area of Habitat Data
 Description: Create Area of Habitat data to characterize species distributions.
     Data are produced following procedures outlined by Brooks et al. (2019)

NEWS.md

@@ -1,3 +1,19 @@
+# aoh 0.0.2.12
+
+- Fix bug in `create_spp_info_data()` in assigning habitat types for resident
+  distributions of migratory and non-migratory species. This bug meant that (i)
+  habitat types for resident distributions of migratory species would include
+  those exclusively affiliated with the species' passage distributions and (ii)
+  habitat types for resident distributions of non-migratory species would not
+  include those exclusively affiliated with the species' passage distributions.
+  Thanks to Jianqiao Zhao for bug report.
+- Update built-in testing data to include multiple migratory species.
+- Update internal R script for creating test dataset
+  (i.e., `inst/scripts/test-data.R`) to be compatible with current version of
+  the package.
+- Update `is_gdal_calc_available()` to be more robust.
+- Fix URLs.
+
 # aoh 0.0.2.11
 
 - Fix bug in `create_spp_info_data()` so that the IUCN Red List API key can be

R/collate_spp_info_data.R

@@ -199,34 +199,63 @@ collate_spp_info_data <- function(x,
   spp_habitat_data <- spp_habitat_data[idx, , drop = FALSE]
   ## if exact matches for habitat codes should not be used then...
   if (isTRUE(adjust_habitat_codes)) {
-    ### create new data by merging habitat codes following KBA guidelines
+    ### create new data by merging habitat codes based on KBA guidelines
+    ### and personal communications with Busana M. and Butchart S.H.M. at
+    ### BirdLife International
     spp_habitat_data <- dplyr::bind_rows(
       list(
         ### resident (non-migratory species)
+        #### NOTE: we include habitat types listed for all seasonal
+        #### distributions. although non-migratory
+        #### species should not have any habitat types affiliated with
+        #### non-residential distributions, this can happen due to
+        #### errors/discrepancies in the IUCN Red List data
         prepare_habitat_codes(
-          dplyr::filter(spp_habitat_data, .data$seasonal %in% migratory_ids),
+          dplyr::filter(spp_habitat_data, !.data$id_no %in% migratory_ids),
           focal_seasonal = 1L,
           extra_seasonal = c(2L, 3L, 4L, 5L, NA_integer_)
         ),
         ### resident (migratory species)
+        #### NOTE: we do not include passage (code 4) habitat types here
+        #### for migratory bird species because these might be substantially
+        #### different from resident distributions
         prepare_habitat_codes(
-          dplyr::filter(spp_habitat_data, !.data$seasonal %in% migratory_ids),
+          dplyr::filter(spp_habitat_data, .data$id_no %in% migratory_ids),
           focal_seasonal = 1L,
           extra_seasonal = c(2L, 3L, 5L, NA_integer_)
         ),
         ### breeding
+        #### NOTE: we include habitat types listed for resident (code 1),
+        #### breeding (code 2) and seasonality uncertain (code 5) when
+        #### mapping the breeding distributions, per KBA guidelines
         prepare_habitat_codes(
           spp_habitat_data,
           focal_seasonal = 2L,
           extra_seasonal = c(1L, 5L, NA_integer_)
         ),
         ### non-breeding
+        #### NOTE: we include habitat types listed for resident (code 1),
+        #### non-breeding (code 2) and seasonality uncertain (code 5) when
+        #### mapping the non-breeding distributions, per KBA guidelines
         prepare_habitat_codes(
           spp_habitat_data,
           focal_seasonal = 3L,
           extra_seasonal = c(1L, 5L, NA_integer_)
         ),
         ### passage
+        #### NOTE: we include habitat types listed for resident (code 1),
+        #### passage (code 24 and seasonality uncertain (code 5) when
+        #### mapping the passage distributions, following a similar
+        #### rationale for the breeding and non-breeding distributions.
+        #### although the resident distribution for migratory species
+        #### does not consider habitat types listed for passage
+        #### distributions (based on the rationale that species might
+        #### have habitat types unique to their passage distribution),
+        #### we include habitat types for resident distributions here
+        #### since (i) many species are missing habitat types for their
+        #### passage distributions and (ii) if the habitat types are
+        #### used during the resident distribution, then they are
+        #### likely used during passage (but not vice versa)
         prepare_habitat_codes(
           spp_habitat_data,
           focal_seasonal = 4L,

R/internal.R

@@ -44,6 +44,9 @@ normalize_path <- function(path, winslash = "\\", mustWork = NA) {
   if (identical(.Platform$OS.type, "unix")) {
     # if on a Unix platform, then we just need to use normalizePath()
     out <- normalizePath(path, winslash = winslash, mustWork = mustWork)
+    # if on a macOS system, this sometimes results in file paths
+    # that contain "//" instead of "/" for a folder separator
+    out <- gsub("//", "/", out, fixed = TRUE)
   } else {
     # if on a Windows platform, then things get a bit more complicated...
     if (!grepl("~", path, fixed = TRUE)) {

R/is_available.R

@@ -138,11 +138,21 @@ is_gdal_calc_available <- function() {
       return(FALSE)
     }
   }
-  v <- try(
-    system(python_gdal_calc("--help"), intern = TRUE),
-    silent = TRUE
+  v <- suppressWarnings(
+    try(
+      system(python_gdal_calc("--help"), intern = TRUE),
+      silent = TRUE
+    )
   )
+  # return FALSE if system command executed with error
   if (inherits(v, "try-error")) return(FALSE)
+  # return FALSE if system command executed with non-zero status
+  if (!is.null(attr(v, "status"))) {
+    if (!identical(attr(v, "status"), 0L)) {
+      return(FALSE)
+    }
+  }
+  # otherwise, return TRUE
   TRUE
 }
 

R/read_spp_range_data.R

@@ -19,7 +19,7 @@ NULL
 #' (see https://www.iucnredlist.org/resources/spatial-data-download).
 #' Data for birds can be obtained by requesting data from
 #' [BirdLife International](https://www.birdlife.org/)
-#' (see <http://datazone.birdlife.org/species/requestdis>).
+#' (see <https://datazone.birdlife.org/species/requestdis>).
 #' To standardize data from the IUCN Red List and BirdLife International,
 #' the `"SISID"` and `"SISRecID"` columns are renamed as `"id_no"`.
 #'

R/terra_gdal_calc.R

@@ -229,6 +229,8 @@ terra_gdal_calc <- function(x, expr,
   if (isTRUE(verbose)) {
     cli::cli_alert_info(paste("System command:", cmd))
   }
+
+  # run gdal_calc via system command
   system(cmd, intern = !verbose)
 
   # clean up
@@ -246,6 +248,15 @@ terra_gdal_calc <- function(x, expr,
     unlink(f3, force = TRUE)
   }
 
+  # verify output file exists
+  assertthat::assert_that(
+    file.exists(filename),
+    msg = paste0(
+      "failed to run `gdal_calc.py` script: output file \"",
+      filename, "\" does not exist"
+    )
+  )
+
   # return result
   if (output_raster) {
     return(stats::setNames(terra::rast(filename), nms))

README.Rmd

@@ -180,7 +180,7 @@ print(spp_aoh_rasters)
 
 Finally, let's create some maps to compare the range data with the Area of habitat data.
 
-```{r "map", message = FALSE, warning = FALSE, dpi = 200, fig.width = 5.5, fig.height = 4, out.width = ifelse(isTRUE(knitr::is_html_output(excludes = c("markdown"))), "60%", "90%")}
+```{r "map", message = FALSE, warning = FALSE, results = "hide", dpi = 200, fig.width = 5.5, fig.height = 4, out.width = ifelse(isTRUE(knitr::is_html_output(excludes = c("markdown"))), "60%", "90%")}
 # create maps
 ## N.B. you might need to install the ggmap package
 map <-

README.md

@@ -366,15 +366,15 @@ produce Area of Habitat data.
     relevant data using:
 
       Hanson JO (2024) aoh: Create Area of Habitat Data. R package version
-      0.0.2.11. Available at https://github.com/prioritizr/aoh.
+      0.0.2.12. Available at https://github.com/prioritizr/aoh.
 
       IUCN [insert year] IUCN Red List of Threatened Species. Version
       [insert version]. Available at www.iucnredlist.org.
 
       BirdLife International and Handbook of the Birds of the World
       ([insert year]) Bird species distribution maps of the world. Version
       [insert version]. Available at
-      http://datazone.birdlife.org/species/requestdis.
+      https://datazone.birdlife.org/species/requestdis.
 
       Jung M, Dahal PR, Butchart SHM, Donald PF, De Lamo X, Lesiv M, Kapos
       V, Rondinini C, Visconti P (2020). "A global map of terrestrial