fix: bug in clusterByDistance() when nrow(tbl) < maxClusters

MazamaScience · Aug 15, 2024 · b835236 · b835236
1 parent a4d2246
commit b835236
Show file tree

Hide file tree

Showing 64 changed files with 125 additions and 77 deletions.
diff --git a/DESCRIPTION b/DESCRIPTION
@@ -1,6 +1,6 @@
 Type: Package
 Package: MazamaLocationUtils
-Version: 0.4.3
+Version: 0.4.4
 Title: Manage Spatial Metadata for Known Locations
 Authors@R: c(
     person("Jonathan", "Callahan", email="[email protected]", role=c("aut","cre")),

diff --git a/NEWS.md b/NEWS.md
@@ -1,3 +1,8 @@
+# MazamaLocationUtils 0.4.4
+
+* Fixed bug in `clusterByDistance()` which could fail when incoming tibbles
+had very few rows.
+
 # MazamaLocationUtils 0.4.3
 
 * Updated `location_getSingleElevation_USGS()` to use new USGS API.

diff --git a/R/clusterByDistance.R b/R/clusterByDistance.R
@@ -128,10 +128,14 @@ clusterByDistance <- function(
 
   # Perform clustering
   for ( clusterCount in 1:maxClusters ) {
-    if ( nrow(tbl) < 2000 ) {
-      clusterObj <- cluster::pam(tbl[,c(lonVar,latVar)],clusterCount)
+
+    if ( nrow(tbl) <= clusterCount ) {
+      # NOTE:  k = positive integer specifying the number of clusters, less than the number of observations.
+      clusterObj <- cluster::pam(tbl[,c(lonVar,latVar)], nrow(tbl) - 1)
+    } else if ( nrow(tbl) < 2000 ) {
+      clusterObj <- cluster::pam(tbl[,c(lonVar,latVar)], clusterCount)
     } else {
-      clusterObj <- cluster::clara(tbl[,c(lonVar,latVar)],clusterCount, samples = 50)
+      clusterObj <- cluster::clara(tbl[,c(lonVar,latVar)], clusterCount, samples = 50)
     }
     clusterLats <- clusterObj$medoids[,latVar]
     diameters <- 2 * clusterObj$clusinfo[,'max_diss'] # decimal degrees
@@ -141,6 +145,19 @@ clusterByDistance <- function(
     radianClusterLats <- clusterLats * pi/180
     meters <- diameters * (1 + cos(radianClusterLats))/2 * 111320
     if ( max(meters) < clusterDiameter ) break
+
+  }
+
+  # NOTE:  If we only had a very small table with widely separate locations,
+  # NOTE:  we can end up never satisfying max(meters) < clusterDiameter. In this
+  # NOTE:  case, every location represents a separate cluster
+  if ( nrow(tbl) <= clusterCount ) {
+    for ( i in 1:seq_len(nrow(tbl)) ) {
+      tbl$clusterLon <- as.numeric(tbl[[lonVar]][i])
+      tbl$clusterLat <- as.numeric(tbl[[latVar]][i])
+      tbl$clusterID <- i
+    }
+    return(tbl)
   }
 
   # Create the vector of deployment identifiers

diff --git a/docs/404.html b/docs/404.html
diff --git a/docs/LICENSE-text.html b/docs/LICENSE-text.html
diff --git a/docs/articles/Developer_Style_Guide.html b/docs/articles/Developer_Style_Guide.html
diff --git a/docs/articles/MazamaLocationUtils.html b/docs/articles/MazamaLocationUtils.html
diff --git a/docs/articles/index.html b/docs/articles/index.html
diff --git a/docs/authors.html b/docs/authors.html
diff --git a/docs/index.html b/docs/index.html
diff --git a/docs/news/index.html b/docs/news/index.html
diff --git a/docs/pkgdown.yml b/docs/pkgdown.yml
@@ -1,8 +1,8 @@
-pandoc: 3.1.2
+pandoc: 3.1.1
 pkgdown: 2.0.7
 pkgdown_sha: ~
 articles:
   Developer_Style_Guide: Developer_Style_Guide.html
   MazamaLocationUtils: MazamaLocationUtils.html
-last_built: 2024-01-23T19:14Z
+last_built: 2024-08-15T02:46Z
 
diff --git a/docs/reference/LocationDataDir.html b/docs/reference/LocationDataDir.html
diff --git a/docs/reference/MazamaLocationUtils.html b/docs/reference/MazamaLocationUtils.html
diff --git a/docs/reference/apiKeys.html b/docs/reference/apiKeys.html
diff --git a/docs/reference/clusterByDistance-1.png b/docs/reference/clusterByDistance-1.png
diff --git a/docs/reference/clusterByDistance.html b/docs/reference/clusterByDistance.html
diff --git a/docs/reference/coreMetadataNames.html b/docs/reference/coreMetadataNames.html
diff --git a/docs/reference/getLocationDataDir.html b/docs/reference/getLocationDataDir.html
diff --git a/docs/reference/id_monitors_500.html b/docs/reference/id_monitors_500.html
diff --git a/docs/reference/index.html b/docs/reference/index.html
diff --git a/docs/reference/initializeMazamaSpatialUtils.html b/docs/reference/initializeMazamaSpatialUtils.html
diff --git a/docs/reference/location_createID.html b/docs/reference/location_createID.html
diff --git a/docs/reference/location_getCensusBlock.html b/docs/reference/location_getCensusBlock.html
diff --git a/docs/reference/location_getOpenCageInfo.html b/docs/reference/location_getOpenCageInfo.html
diff --git a/docs/reference/location_getSingleAddress_Photon.html b/docs/reference/location_getSingleAddress_Photon.html
diff --git a/docs/reference/location_getSingleAddress_TexasAM.html b/docs/reference/location_getSingleAddress_TexasAM.html
diff --git a/docs/reference/location_getSingleElevation_USGS.html b/docs/reference/location_getSingleElevation_USGS.html
diff --git a/docs/reference/location_initialize.html b/docs/reference/location_initialize.html