Added saving and reading nn-data to/from JSON-file

.gitignore

@@ -1,4 +1,5 @@
 .idea
 target
 out
-artifacts
+artifacts
+nn_data.json

README.md

@@ -12,32 +12,39 @@ If you want to learn more about Neural Networks check out these YouTube-playlist
 
 - Neural Network with variable amounts of inputs, hidden nodes and outputs
 - Two layered (hidden + output)
-- [Maven](https://maven.apache.org) Build-Management
+- Save the weights and biases of a NN to a JSON-file
+- Read a JSON-file with NN-data
 - [EJML](https://www.ejml.org) used for Matrix math
-
-## Examples (Usages)
-
-- [XOR solved with Basic Neural Network Library](https://github.com/kim-marcel/xor_with_nn)
-
-## Download
-
-If you want to use this library you can [download](https://github.com/kim-marcel/basic_neural_network/releases/download/v0.1-alpha/basic_neural_network-v0.1-alpha.jar) v0.1-alpha here or check the release tab of this repository. There might be a newer version available.
+- [Maven](https://maven.apache.org) Build-Management
 
 ## Code example
 
 ```java
 // Neural Network with 2 inputs, 4 hidden nodes and 1 output
 NeuralNetwork nn = new NeuralNetwork(2, 4, 1);
 
+// Reads from a (previously generated) JSON-file the weights and biases of the NN
+nn.readFromFile();
+
 // Train the Neural Network with a training dataset
 nn.train(trainingDataInputs, trainingDataTargets);
 
 // Guess for the given testing data is returned as a 2D array (double[][])
 nn.guess(testingData);
+
+// Writes a JSON-file with the current "state" (weights and biases) of the NN
+nn.writeToFile();
 ```
-For a more detailed example check out [this](https://github.com/kim-marcel/xor_with_nn) repository.
+A more detailed example cam be found below.
+
+## Download
+
+If you want to use this library you can download [v0.1-alpha](https://github.com/kim-marcel/basic_neural_network/releases/download/v0.1-alpha/basic_neural_network-v0.1-alpha.jar) here or check the release tab of this repository.
+
+## Examples
+
+- [XOR solved with Basic Neural Network Library](https://github.com/kim-marcel/xor_with_nn)
 
 ## Upcoming features
 
 - Support for multiple layers
-- Save and load the neural network to a file

pom.xml

@@ -8,6 +8,19 @@
     <artifactId>basic_neural_network</artifactId>
     <version>v0.1-alpha</version>
 
+    <build>
+        <plugins>
+            <plugin>
+                <groupId>org.apache.maven.plugins</groupId>
+                <artifactId>maven-compiler-plugin</artifactId>
+                <configuration>
+                    <source>1.8</source>
+                    <target>1.8</target>
+                </configuration>
+            </plugin>
+        </plugins>
+    </build>
+
     <dependencies>
 
         <dependency>
@@ -16,6 +29,12 @@
             <version>0.33</version>
         </dependency>
 
+        <dependency>
+            <groupId>com.googlecode.json-simple</groupId>
+            <artifactId>json-simple</artifactId>
+            <version>1.1.1</version>
+        </dependency>
+
     </dependencies>
 
 

src/main/java/NeuralNetwork.java

@@ -1,7 +1,11 @@
 import org.ejml.simple.SimpleMatrix;
+import org.json.simple.JSONObject;
+import org.json.simple.parser.JSONParser;
+import org.json.simple.parser.ParseException;
+import utilities.MatrixConverter;
 import utilities.Sigmoid;
 
-import java.util.Arrays;
+import java.io.*;
 import java.util.Random;
 
 /**
@@ -21,11 +25,11 @@ public class NeuralNetwork {
     private SimpleMatrix weightsIH; // matrix with weights between input and hidden layer
     private SimpleMatrix weightsHO; // matrix with weights between hidden and output layer
 
-    private SimpleMatrix biasH;
-    private SimpleMatrix biasO;
+    private SimpleMatrix biasH; // bias of the hidden layer
+    private SimpleMatrix biasO; // bias of the output layer
 
     // Constructor
-    // generate a new neural network with 1 hidden layer with the given amount of nodes in the layers
+    // generate a new neural network with 1 hidden layer with the given amount of nodes in the individual layers
     public NeuralNetwork(int inputNodes, int hiddenNodes, int outputNodes){
         this.inputNodes = inputNodes;
         this.hiddenNodes = hiddenNodes;
@@ -43,17 +47,17 @@ public NeuralNetwork(int inputNodes, int hiddenNodes, int outputNodes){
     // guess method, input is a one column matrix with the input values
     public double[][] guess(double[] i){
         // transform array to matrix
-        SimpleMatrix inputs = arrayToMatrix(i);
+        SimpleMatrix inputs = MatrixConverter.arrayToMatrix(i);
 
         SimpleMatrix hidden = calculateLayer(weightsIH, biasH, inputs);
         SimpleMatrix output = calculateLayer(weightsHO, biasO, hidden);
-        return matrixToArray(output);
+        return MatrixConverter.matrixToArray(output);
     }
 
     public void train(double[] i, double[] t){
         // transform 2d array to matrix
-        SimpleMatrix inputs = arrayToMatrix(i);
-        SimpleMatrix targets = arrayToMatrix(t);
+        SimpleMatrix inputs = MatrixConverter.arrayToMatrix(i);
+        SimpleMatrix targets = MatrixConverter.arrayToMatrix(t);
 
         // calculate outputs of hidden and output layer for the given inputs
         SimpleMatrix hidden = calculateLayer(weightsIH, biasH, inputs);
@@ -100,19 +104,52 @@ private SimpleMatrix calculateDeltas(SimpleMatrix gradient, SimpleMatrix layer){
         return gradient.mult(layer.transpose());
     }
 
-    private SimpleMatrix arrayToMatrix(double[] i){
-        double[][] input = {i};
-        return new SimpleMatrix(input).transpose();
+    public void writeToFile(){
+        JSONObject obj = new JSONObject();
+        obj.put("weightsIH", MatrixConverter.matrixToJSON(weightsIH));
+        obj.put("weightsHO", MatrixConverter.matrixToJSON(weightsHO));
+        obj.put("biasH", MatrixConverter.matrixToJSON(biasH));
+        obj.put("biasO", MatrixConverter.matrixToJSON(biasO));
+
+        try (FileWriter file = new FileWriter("nn_data.json")) {
+
+            file.write(obj.toJSONString());
+            file.flush();
+
+        } catch (IOException e) {
+            e.printStackTrace();
+        }
+
     }
 
-    public double[][] matrixToArray(SimpleMatrix i){
-        double[][] result = new double[i.numRows()][i.numCols()];
-        for (int j = 0; j < result.length; j++) {
-            for (int k = 0; k < result[0].length; k++) {
-                result[j][k] = i.get(j, k);
-            }
+    public void readFromFile(){
+        JSONParser parser = new JSONParser();
+
+        try {
+
+            Object obj = parser.parse(new FileReader("nn_data.json"));
+
+            JSONObject nnData = (JSONObject) obj;
+
+            JSONObject weightsIHJSON = (JSONObject) nnData.get("weightsIH");
+            JSONObject weightsHOJSON = (JSONObject) nnData.get("weightsHO");
+
+            weightsIH = MatrixConverter.jsonToMatrix(weightsIHJSON);
+            weightsHO = MatrixConverter.jsonToMatrix(weightsHOJSON);
+
+            JSONObject biasHJSON = (JSONObject) nnData.get("biasH");
+            JSONObject biasOJSON = (JSONObject) nnData.get("biasO");
+
+            biasH = MatrixConverter.jsonToMatrix(biasHJSON);
+            biasO = MatrixConverter.jsonToMatrix(biasOJSON);
+
+        } catch (FileNotFoundException e) {
+            e.printStackTrace();
+        } catch (IOException e) {
+            e.printStackTrace();
+        } catch (ParseException e) {
+            e.printStackTrace();
         }
-        return result;
     }
 
     public double getLearningRate() {

src/main/java/utilities/MatrixConverter.java

@@ -0,0 +1,57 @@
+package utilities;
+
+import org.ejml.simple.SimpleMatrix;
+import org.json.simple.JSONObject;
+
+/**
+ * Created by KimFeichtinger on 07.03.18.
+ */
+public class MatrixConverter {
+
+    public static SimpleMatrix arrayToMatrix(double[] i){
+        double[][] input = {i};
+        return new SimpleMatrix(input).transpose();
+    }
+
+    public static double[][] matrixToArray(SimpleMatrix i){
+        double[][] result = new double[i.numRows()][i.numCols()];
+        for (int j = 0; j < result.length; j++) {
+            for (int k = 0; k < result[0].length; k++) {
+                result[j][k] = i.get(j, k);
+            }
+        }
+        return result;
+    }
+
+    public static SimpleMatrix jsonToMatrix(JSONObject i){
+        int rows = i.size();
+        int cols = ((JSONObject) i.get(Integer.toString(0))).size();
+
+        SimpleMatrix result = new SimpleMatrix(rows, cols);
+
+        for (int j = 0; j < i.size(); j++) {
+            JSONObject js = (JSONObject) i.get(Integer.toString(j));
+            for (int k = 0; k < js.size(); k++) {
+                double d = (double) js.get(Integer.toString(k));
+                result.set(j, k, d);
+            }
+        }
+
+        return result;
+    }
+
+    public static JSONObject matrixToJSON(SimpleMatrix i){
+        JSONObject result = new JSONObject();
+        JSONObject a = new JSONObject();
+
+        for (int j = 0; j < i.numRows(); j++) {
+            a.clear();
+            for (int k = 0; k < i.numCols(); k++) {
+                a.put(k, i.get(j, k));
+            }
+            result.put(j, new JSONObject(a));
+        }
+
+        return result;
+    }
+}