Support gather of pytorch

api/src/main/java/ai/djl/ndarray/NDArray.java

@@ -545,6 +545,17 @@ default NDArray get(NDArray index) {
         return get(new NDIndex().addBooleanIndex(index));
     }
 
+    /**
+     * Returns a partial {@code NDArray} pointed by the indexed array. Given NDArray arr, NDArray
+     * idx, and long axis, the output is out_{ijk} = arr_{idx_{ijk}, j, k} if axis=0 or arr_{i,
+     * idx_{ijk}, k} if axis=1 or arr_{i, j, idx_{ijk}} if axis=2
+     *
+     * @param index picks the elements of an NDArray to the same position as index
+     * @param axis the entries of index are indices of axis
+     * @return the partial {@code NDArray} of the same shape as index
+     */
+    NDArray gather(NDArray index, int axis);
+
     /**
      * Returns a scalar {@code NDArray} corresponding to a single element.
      *

api/src/main/java/ai/djl/ndarray/NDArrayAdapter.java

@@ -176,6 +176,12 @@ public String[] toStringArray(Charset charset) {
         throw new UnsupportedOperationException(UNSUPPORTED_MSG);
     }
 
+    /** {@inheritDoc} */
+    @Override
+    public NDArray gather(NDArray index, int axis) {
+        throw new UnsupportedOperationException(UNSUPPORTED_MSG);
+    }
+
     /** {@inheritDoc} */
     @Override
     public void set(Buffer data) {

docs/README.md

@@ -4,6 +4,8 @@ This folder contains examples and documentation for the Deep Java Library (DJL)
 
 ## [JavaDoc API Reference](https://javadoc.djl.ai/)
 
+Note: when searching in JavaDoc, if your access is denied, please try removing the string `undefined` in the url.
+
 ## [Demos](https://djl.ai/website/demo.html)
 
 ## Cheat sheet

docs/get.md

@@ -90,9 +90,19 @@ If you build following the above instructions, you will use the version of the c
 
 You can look here to find the [list of DJL releases](https://github.com/deepjavalibrary/djl/releases).
 
-### Using build from source in another project
+### Using built-from-source version in another project
 
-If you have another project and want to use a custom version of DJL in it, you can do this by building from source. Once you have built DJL from source, run `./gradlew publishToMavenLocal`. This will install DJL to your local maven repository cache located on your filesystem at `~/.m2/repository`. When you publish here, it will use the same snapshot versions as used by the snapshot repository above for adding the DJL dependencies.
+If you have another project and want to use a custom version of DJL in it, you can do the following. First, build DJL from source by running `./gradlew build` inside djl folder. Then run `./gradlew publishToMavenLocal`, which will install DJL to your local maven repository cache, located on your filesystem at `~/.m2/repository`. After publishing it here, you can add the DJL snapshot version dependencies as shown below 
+
+```groovy
+dependencies {
+    implementation platform("ai.djl:bom:0.17.0-SNAPSHOT")
+}
+```
+
+This snapshot version is the same as the custom DJL repository. 
+
+You also need to change directory to `djl/bom`. Then build and publish it to maven local same as was done in `djl`.
 
 From there, you may have to update the Maven or Gradle build of the project importing DJL to also look at the local maven repository cache for your locally published versions of DJL. For Maven, no changes are necessary. If you are using Gradle, you will have to add the maven local repository such as this [example](https://github.com/deepjavalibrary/djl-demo/blob/135c969d66d98d1672852e53a37e52ca1da3e325/pneumonia-detection/build.gradle#L11):
 
@@ -101,3 +111,5 @@ repositories {
     mavenLocal()
 }
 ```
+
+Note that `mavenCentral()` may still be needed for applications like log4j and json.

engines/mxnet/mxnet-engine/src/main/java/ai/djl/mxnet/engine/MxNDArray.java

@@ -306,6 +306,12 @@ public void set(Buffer data) {
         JnaUtils.syncCopyFromCPU(getHandle(), buf, size);
     }
 
+    /** {@inheritDoc} */
+    @Override
+    public NDArray gather(NDArray index, int axis) {
+        throw new UnsupportedOperationException("Not implemented yet.");
+    }
+
     /** {@inheritDoc} */
     @Override
     public void copyTo(NDArray ndArray) {

engines/pytorch/pytorch-engine/src/main/java/ai/djl/pytorch/engine/PtNDArray.java

@@ -239,6 +239,15 @@ public NDArray get(long... indices) {
         return JniUtils.getItem(this, indices);
     }
 
+    /** {@inheritDoc} */
+    @Override
+    public NDArray gather(NDArray index, int axis) {
+        if (!(index instanceof PtNDArray)) {
+            throw new IllegalArgumentException("Only PtNDArray is supported.");
+        }
+        return JniUtils.gather(this, (PtNDArray) index, axis);
+    }
+
     /** {@inheritDoc} */
     @Override
     public void copyTo(NDArray array) {

engines/pytorch/pytorch-engine/src/main/java/ai/djl/pytorch/jni/JniUtils.java

@@ -346,6 +346,12 @@ public static void set(PtNDArray self, ByteBuffer data) {
         PyTorchLibrary.LIB.torchSet(self.getHandle(), data);
     }
 
+    public static PtNDArray gather(PtNDArray ndArray, PtNDArray index, long dim) {
+        return new PtNDArray(
+                ndArray.getManager(),
+                PyTorchLibrary.LIB.torchGather(ndArray.getHandle(), index.getHandle(), dim, false));
+    }
+
     public static PtNDArray pick(PtNDArray ndArray, PtNDArray index, long dim) {
         Shape indexShape = index.getShape();
         Shape ndShape = ndArray.getShape();

engines/tensorflow/tensorflow-engine/src/main/java/ai/djl/tensorflow/engine/TfNDArray.java

@@ -206,6 +206,12 @@ public void set(Buffer data) {
         JavacppUtils.setByteBuffer(getHandle(), buf);
     }
 
+    /** {@inheritDoc} */
+    @Override
+    public NDArray gather(NDArray index, int axis) {
+        throw new UnsupportedOperationException("Not implemented yet.");
+    }
+
     /** {@inheritDoc} */
     @Override
     public void attach(NDManager manager) {

gradle/wrapper/gradle-wrapper.properties

@@ -3,4 +3,4 @@ distributionBase=GRADLE_USER_HOME
 distributionPath=wrapper/dists
 zipStoreBase=GRADLE_USER_HOME
 zipStorePath=wrapper/dists
-distributionUrl=https\://services.gradle.org/distributions/gradle-7.4.1-bin.zip
+distributionUrl=https\://services.gradle.org/distributions/gradle-7.4.2-bin.zip

integration/src/main/java/ai/djl/integration/tests/ndarray/NDIndexTest.java

@@ -12,10 +12,12 @@
  */
 package ai.djl.integration.tests.ndarray;
 
+import ai.djl.engine.Engine;
 import ai.djl.ndarray.NDArray;
 import ai.djl.ndarray.NDManager;
 import ai.djl.ndarray.index.NDIndex;
 import ai.djl.ndarray.types.Shape;
+import ai.djl.testing.TestRequirements;
 import org.testng.Assert;
 import org.testng.annotations.Test;
 
@@ -52,6 +54,23 @@ public void testPick() {
         }
     }
 
+    @Test
+    public void testGather() {
+        // Currently in windows gradle cannot find all the engines.
+        // In the dependency, changing runtimeOnly to api however will remedy the problem.
+        // TODO: remove this when gradle problem is fixed.
+        TestRequirements.notWindows();
+        Engine engine = Engine.getEngine("PyTorch");
+        try (NDManager manager = engine.newBaseManager()) {
+            NDArray arr = manager.arange(20f).reshape(-1, 4);
+            long[] idx = {0, 0, 2, 1, 1, 2};
+            NDArray index = manager.create(idx, new Shape(3, 2));
+            NDArray actual = arr.gather(index, 1);
+            NDArray expected = manager.create(new float[] {0, 0, 6, 5, 9, 10}, new Shape(3, 2));
+            Assert.assertEquals(actual, expected);
+        }
+    }
+
     @Test
     public void testGet() {
         try (NDManager manager = NDManager.newBaseManager()) {