#6058 TSNE: throw useful exception for no points (distance may be undefined)

deeplearning4j/deeplearning4j-manifold/deeplearning4j-tsne/src/main/java/org/deeplearning4j/plot/BarnesHutTsne.java

@@ -242,6 +242,12 @@ public INDArray computeGaussianPerplexity(final INDArray d, double u) {
                 tree.search(d.slice(i), k + 1, results, new ArrayList<Double>());
                 double betas = beta.getDouble(i);
 
+                if(results.size() == 0){
+                    throw new IllegalStateException("Search returned no values for vector " + i +
+                            " - similarity \"" + simiarlityFunction + "\" may not be defined (for example, vector is" +
+                            " all zeros with cosine similarity)");
+                }
+
                 INDArray cArr = VPTree.buildFromData(results);
                 Pair<INDArray, Double> pair = computeGaussianKernel(cArr, beta.getDouble(i), k);
                 INDArray currP = pair.getFirst();

deeplearning4j/deeplearning4j-manifold/deeplearning4j-tsne/src/test/java/org/deeplearning4j/plot/Test6058.java

@@ -0,0 +1,47 @@
+package org.deeplearning4j.plot;
+
+import lombok.val;
+import org.junit.Test;
+import org.nd4j.linalg.api.ndarray.INDArray;
+import org.nd4j.linalg.factory.Nd4j;
+
+import java.util.ArrayList;
+
+import static org.junit.Assert.assertTrue;
+
+public class Test6058 {
+
+    @Test
+    public void test() throws Exception {
+        //All zero input -> cosine similarity isn't defined
+        //https://github.com/deeplearning4j/deeplearning4j/issues/6058
+        val iterations = 10;
+        val cacheList = new ArrayList<String>();
+
+        int nWords  = 100;
+        for(int i=0; i<nWords; i++ ) {
+            cacheList.add("word_" + i);
+        }
+
+        //STEP 3: build a dual-tree tsne to use later
+        System.out.println("Build model....");
+        val tsne = new BarnesHutTsne.Builder()
+                .setMaxIter(iterations)
+                .theta(0.5)
+                .normalize(false)
+                .learningRate(1000)
+                .useAdaGrad(false)
+                //.usePca(false)
+                .build();
+
+        System.out.println("fit");
+        INDArray weights = Nd4j.rand(new int[]{nWords, 100});
+        weights.getRow(1).assign(0);
+        try {
+            tsne.fit(weights);
+        } catch (IllegalStateException e){
+            assertTrue(e.getMessage().contains("may not be defined"));
+        }
+    }
+
+}