Merge pull request #25 from JuliaAI/clustering-0-15-compat

Bump compat for Clustering to 0.15

Author: ablaom

Project.toml

@@ -1,15 +1,15 @@
 name = "MLJClusteringInterface"
 uuid = "d354fa79-ed1c-40d4-88ef-b8c7bd1568af"
 authors = ["Anthony D. Blaom <[email protected]>", "Thibaut Lienart <[email protected]>", "Okon Samuel <[email protected]>"]
-version = "0.1.9"
+version = "0.1.10"
 
 [deps]
 Clustering = "aaaa29a8-35af-508c-8bc3-b662a17a0fe5"
 Distances = "b4f34e82-e78d-54a5-968a-f98e89d6e8f7"
 MLJModelInterface = "e80e1ace-859a-464e-9ed9-23947d8ae3ea"
 
 [compat]
-Clustering = "0.14"
+Clustering = "0.15"
 Distances = "0.9, 0.10"
 MLJModelInterface = "1.4"
 julia = "1.6"

src/MLJClusteringInterface.jl

@@ -139,7 +139,7 @@ function MMI.predict(model::DBSCAN, ::Nothing, X)
         leafsize=model.leafsize,
         min_neighbors=model.min_neighbors,
         min_cluster_size=model.min_cluster_size,
-    )
+    ).clusters
     nclusters = length(clusters)
 
     # assignments and point types

test/Project.toml

@@ -5,4 +5,5 @@ LinearAlgebra = "37e2e46d-f89d-539d-b4ee-838fcccc9c8e"
 MLJBase = "a7f614a8-145f-11e9-1d2a-a57a1082229d"
 MLJTestInterface = "72560011-54dd-4dc2-94f3-c5de45b75ecd"
 Random = "9a3f8284-a2c9-5f02-9a11-845980a1fd5c"
+StableRNGs = "860ef19b-820b-49d6-a774-d7a799459cd3"
 Test = "8dfed614-e22c-5e08-85e1-65c5234f0b40"

test/runtests.jl

@@ -5,10 +5,11 @@ import LinearAlgebra: norm
 using MLJBase
 using MLJTestInterface
 using MLJClusteringInterface
-using Random: seed!
+using StableRNGs
+using Random
 using Test
 
-seed!(132442)
+srng(n) = StableRNGs.StableRNG(n)
 X, y = @load_crabs
 
 
@@ -59,6 +60,35 @@ end
 
 @testset "DBSCAN" begin
 
+    # test interface is faithful on some synthetic data:
+    X, _ = make_moons(500, rng=srng(123))
+    dbscan = DBSCAN(radius=0.15, min_neighbors=3)
+    yhat3, _report = predict(dbscan, nothing, X)
+    Xarray = MLJBase.matrix(X)'
+    clusters = Clustering.dbscan(
+        Xarray,
+        dbscan.radius;
+        leafsize=dbscan.leafsize,
+        min_neighbors=dbscan.min_neighbors,
+        min_cluster_size=dbscan.min_cluster_size,
+    ).clusters
+    noisy_indices = Set(1:length(y)) # initialization
+    for (k, cluster) in enumerate(clusters)
+        for i in cluster.core_indices
+            delete!(noisy_indices, i)
+            @test _report.point_types[i] == 'C'
+            @test yhat3[i] == k
+        end
+        for i in cluster.boundary_indices
+            delete!(noisy_indices, i)
+            @test _report.point_types[i] == 'B'
+            @test yhat3[i] == k
+        end
+    end
+    @test all(noisy_indices) do i
+        _report.point_types[i] == 'N' && yhat3[i] == 0
+    end
+
     # five spot pattern
     X = [
         0.0 0.0
@@ -72,7 +102,7 @@ end
     dbscan = DBSCAN(radius=0.1)
     yhat1, report1 = predict(dbscan, nothing, X)
     @test report1.nclusters == 5
-    @test report1.point_types == fill('B', 5)
+    @test report1.point_types == fill('C', 5)
     @test Set(yhat1) == Set(unique(yhat1))
     @test Set(report1.cluster_labels) == Set(unique(yhat1))