Remove AsyncTask

Signed-off-by: Paul Dubs <[email protected]>

Author: RobAltena

android-examples/app/src/main/java/com/example/androidDl4jClassifier/MainActivity.java

@@ -23,6 +23,7 @@ public boolean onOptionsItemSelected(MenuItem item) {
         switch (item.getItemId()) {
             case R.id.action_linear:
                 // User chose linear dataset.
+                ScatterView view = findViewById(R.id.id_scatterview);
                 return true;
 
             case R.id.action_moon:

android-examples/app/src/main/java/com/example/androidDl4jClassifier/ScatterView.java

@@ -4,49 +4,27 @@
 import android.graphics.Canvas;
 import android.graphics.Color;
 import android.graphics.Paint;
-import android.os.AsyncTask;
 import android.util.AttributeSet;
 import android.view.View;
 
 import androidx.annotation.Nullable;
 
-import org.deeplearning4j.nn.conf.MultiLayerConfiguration;
-import org.deeplearning4j.nn.conf.NeuralNetConfiguration;
-import org.deeplearning4j.nn.conf.layers.DenseLayer;
-import org.deeplearning4j.nn.conf.layers.OutputLayer;
-import org.deeplearning4j.nn.multilayer.MultiLayerNetwork;
-import org.deeplearning4j.nn.weights.WeightInit;
-import org.deeplearning4j.optimize.listeners.ScoreIterationListener;
-import org.nd4j.evaluation.classification.Evaluation;
-import org.nd4j.linalg.activations.Activation;
-import org.nd4j.linalg.api.buffer.DataType;
 import org.nd4j.linalg.api.ndarray.INDArray;
-import org.nd4j.linalg.dataset.DataSet;
-import org.nd4j.linalg.factory.Nd4j;
-import org.nd4j.linalg.learning.config.Nesterovs;
-import org.nd4j.linalg.lossfunctions.LossFunctions;
 
-import java.io.BufferedReader;
-import java.io.IOException;
-import java.io.InputStreamReader;
-import java.util.ArrayList;
-
-public class ScatterView extends View {
+public class ScatterView extends View implements OnTrainingUpdateEventListener{
 
     private final Paint redPaint;
     private final Paint greenPaint;
     private final Paint lightGreenPaint;
     private final Paint lightRedPaint;
-    private float[][] data;
-    private DataSet ds;
 
-    private final int nPointsPerAxis = 100;
-    private INDArray xyGrid; //x,y grid to calculate the output image. Needs to be calculated once, then re-used.
-    private INDArray modelOut = null;
+
+    private INDArray modelOut = null; // nn output for the grid.
+
+    private TrainingTask task;
 
     public ScatterView(Context context, @Nullable AttributeSet attrs) {
         super(context, attrs);
-        data = null;
         redPaint = new Paint();
         redPaint.setColor(Color.RED);
         greenPaint = new Paint();
@@ -57,16 +35,9 @@ public ScatterView(Context context, @Nullable AttributeSet attrs) {
         lightRedPaint = new Paint();
         lightRedPaint.setColor(Color.rgb(255, 153, 152));
 
-        AsyncTask.execute(() -> {
-            try {
-                calcGrid();
-                ReadCSV();
-                BuildNN();
-
-            } catch (IOException e) {
-                e.printStackTrace();
-            }
-        });
+        task = new TrainingTask();
+        task.setListener(this);
+        task.executeTask();
     }
 
     @Override
@@ -75,9 +46,10 @@ public void onDraw(Canvas canvas) {
         int w = this.getWidth();
 
         //draw the nn predictions:
-        if ((modelOut != null) && (null != xyGrid )){
-            int halfRectHeight = h / nPointsPerAxis;
-            int halfRectWidth = w / nPointsPerAxis;
+        if (modelOut != null) {
+            int halfRectHeight = h / task.getnPointsPerAxis();
+            int halfRectWidth = w / task.getnPointsPerAxis();
+            INDArray  xyGrid = task.getXyGrid();
             int nRows = xyGrid.rows();
 
             for (int i = 0; i< nRows; i++){
@@ -91,9 +63,9 @@ public void onDraw(Canvas canvas) {
         }
 
         //draw the data set if we have one.
-        if (null != data) {
+        if (null != task.getData()) {
 
-            for (float[] datum : data) {
+            for (float[] datum : task.getData()) {
                 int x = (int) (datum[1] * w);
                 int y = (int) (datum[2] * h);
                 Paint p = (datum[0] == 0.0f) ? redPaint : greenPaint;
@@ -102,124 +74,11 @@ public void onDraw(Canvas canvas) {
         }
     }
 
-    /**
-     * this is not the regular way to read a csv file into a data set with DL4j.
-     * In this example we have put the data in the assets folder so that the demo works offline.
-     */
-    private void ReadCSV() throws IOException {
-        InputStreamReader is = new InputStreamReader(MainActivity.getInstance().getApplicationContext().getAssets()
-                .open("linear_data_train.csv"));
-
-        BufferedReader reader = new BufferedReader(is);
-        ArrayList<String> rawSVC = new ArrayList<>();
-        String line;
-        while ((line = reader.readLine()) != null) {
-            rawSVC.add(line);
-        }
-
-        float[][] tmpData = new float[rawSVC.size()][3];
-
-        int index = 0;
-        for(String l : rawSVC){
-            String[] values = l.split(",");
-            for(int col = 0; col< 3L; col++){
-                tmpData[index][col] = Float.parseFloat(values[col]);
-            }
-
-            index++;
-        }
-
-        normalizeColumn(1, tmpData);
-        normalizeColumn(2, tmpData);
-
-        this.data = tmpData;
-        INDArray arrData = Nd4j.createFromArray(tmpData);
-        INDArray arrFeatures = arrData.getColumns(1, 2);
-        INDArray c1 = arrData.getColumns(0);
-        INDArray c2 = c1.mul(-1).addi(1.0);
-        INDArray labels = Nd4j.hstack(c1, c2);
-        ds = new DataSet(arrFeatures, labels);
-    }
-
-    /**
-     *  Normalize the data in a given column. Normally one would use datavec.
-     * @param c column to normalise.
-     * @param tmpData java float array.
-     */
-    private void normalizeColumn(int c, float[][] tmpData){
-        int numPoints = tmpData.length;
-        float min= tmpData[0][c];
-        float max= tmpData[0][c];
-        for (float[] tmpDatum : tmpData) {
-            float x = tmpDatum[c];
-            if (x < min) {
-                min = x;
-            }
-            if (x > max) {
-                max = x;
-            }
-        }
-
-        for (int i=0; i<numPoints; i++){
-            float x = tmpData[i][c];
-            tmpData[i][c] = (x - min)  / (max - min);
-        }
-    }
-
-    private void BuildNN(){
-        int seed = 123;
-        double learningRate = 0.005;
-        int numInputs = 2;
-        int numOutputs = 2;
-        int numHiddenNodes = 20;
-        int nEpochs = 2000;
-
-        MultiLayerConfiguration conf = new NeuralNetConfiguration.Builder()
-                .seed(seed)
-                .weightInit(WeightInit.XAVIER)
-                .updater(new Nesterovs(learningRate, 0.9))
-                .list()
-                .layer(new DenseLayer.Builder().nIn(numInputs).nOut(numHiddenNodes)
-                        .activation(Activation.RELU)
-                        .build())
-                .layer(new OutputLayer.Builder(LossFunctions.LossFunction.NEGATIVELOGLIKELIHOOD)
-                        .activation(Activation.SOFTMAX)
-                        .nIn(numHiddenNodes).nOut(numOutputs).build())
-                .build();
-
-        MultiLayerNetwork model = new MultiLayerNetwork(conf);
-        model.init();
-        model.setListeners(new ScoreIterationListener(10));
-
-        for(int i = 0; i<nEpochs; i++){
-            model.fit(ds);
-            INDArray tmp = model.output(xyGrid);
-
-            this.post(() -> {
-                this.modelOut =  tmp; // update from within the UI thread.
-                this.invalidate(); // have the UI thread redraw at its own convenience.
-            });
-        }
-
-        Evaluation eval = new Evaluation(numOutputs);
-        INDArray features = ds.getFeatures();
-        INDArray labels = ds.getLabels();
-        INDArray predicted = model.output(features,false);
-        eval.eval(labels, predicted);
-        System.out.println(eval.stats());
-
-        this.invalidate();
-    }
-    /**
-     * The x,y grid to calculate the NN output. Only needs to be calculated once.
-     */
-    private void calcGrid(){
-        // x coordinates of the pixels for the NN.
-        INDArray xPixels = Nd4j.linspace(0, 1.0, nPointsPerAxis, DataType.DOUBLE);
-        // y coordinates of the pixels for the NN.
-        INDArray yPixels = Nd4j.linspace(0, 1.0, nPointsPerAxis, DataType.DOUBLE);
-        //create the mesh:
-        INDArray [] mesh = Nd4j.meshgrid(xPixels, yPixels);
-        xyGrid = Nd4j.vstack(mesh[0].ravel(), mesh[1].ravel()).transpose();
+    @Override
+    public void OnTrainingUpdate(INDArray modelOut) {
+        this.post(() -> {
+            this.modelOut =  modelOut; // update from within the UI thread.
+            this.invalidate(); // have the UI thread redraw at its own convenience.
+        });
     }
 }