Added generateTrainingData() function

Author: generic-github-user

script.js

@@ -123,47 +123,51 @@ for (var i = 0; i < numTrainingImages; i ++) {
 
 // Wait for last image (testing data) to load before continuing
 trainingData.images[trainingData.images.length - 1].onload = function () {
-// Create training data from pixels of image elements
-// Create a new variable to store the data
-var pixels;
-var pixelsArray;
-var outputValues;
-// Loop through each training image
-
-for (var i = 0; i < numTrainingImages; i ++) {
-// Create a tensor with 3 (RGB) color channels from the image element
-pixels = tf.fromPixels(trainingData.images[i], numParameters);
-// Resize image to the specified dimensions with resizeBilinear()
-pixels = tf.image.resizeBilinear(pixels, [imageSize, imageSize]);
-// Get the values array from the pixels tensor
-pixels = pixels.dataSync();
-// Add new array to trainingData.pixels.input to store the pixel values for the image
-pixelsArray = [];
-// Loop through each value in the pixels array
-// The whole pixels array is not pushed on at once because the array format will be incompatible
-pixels.forEach(
-// Add color value to the corresponding image's trainingData.pixels.input array
-(element) => pixelsArray.push(element)
-);
-trainingData.pixels.input.push(pixelsArray);
-outputValues = new Array(numParameters).fill(1);
-trainingData.pixels.output.push(outputValues);
-
-// Uncaught Error: Constructing tensor of shape (92160) should match the length of values (46095)
-const generated = generator.model.predict(parameters.display).dataSync();
-const generatedArray = [];
-generated.forEach(
-(element) => generatedArray.push(element)
-);
-trainingData.pixels.input.push(generatedArray);
+function generateTrainingData() {
+// Create training data from pixels of image elements
+// Create a new variable to store the data
+var pixels;
+var pixelsArray;
+var outputValues;
+// Loop through each training image
+
+for (var i = 0; i < numTrainingImages; i ++) {
+// Create a tensor with 3 (RGB) color channels from the image element
+pixels = tf.fromPixels(trainingData.images[i], numParameters);
+// Resize image to the specified dimensions with resizeBilinear()
+pixels = tf.image.resizeBilinear(pixels, [imageSize, imageSize]);
+// Get the values array from the pixels tensor
+pixels = pixels.dataSync();
+// Add new array to trainingData.pixels.input to store the pixel values for the image
+pixelsArray = [];
+// Loop through each value in the pixels array
+// The whole pixels array is not pushed on at once because the array format will be incompatible
+pixels.forEach(
+// Add color value to the corresponding image's trainingData.pixels.input array
+(element) => pixelsArray.push(element)
+);
+trainingData.pixels.input.push(pixelsArray);
+outputValues = new Array(numParameters).fill(1);
+trainingData.pixels.output.push(outputValues);
+
+// Uncaught Error: Constructing tensor of shape (92160) should match the length of values (46095)
+const generated = generator.model.predict(parameters.display).dataSync();
+const generatedArray = [];
+generated.forEach(
+(element) => generatedArray.push(element)
+);
+trainingData.pixels.input.push(generatedArray);
+
+outputValues = new Array(numParameters).fill(-1);
+trainingData.pixels.output.push(outputValues);
+}
 
-outputValues = new Array(numParameters).fill(-1);
-trainingData.pixels.output.push(outputValues);
+// Create a tensor from the pixel values of the training data and store it in trainingData.tensor.input
+trainingData.tensor.input = tf.tensor(trainingData.pixels.input);
+trainingData.tensor.output = tf.tensor(trainingData.pixels.output);
 }
 
-// Create a tensor from the pixel values of the training data and store it in trainingData.tensor.input
-trainingData.tensor.input = tf.tensor(trainingData.pixels.input);
-trainingData.tensor.output = tf.tensor(trainingData.pixels.output);
+generateTrainingData();
 
 // trainingData.tensor.output = tf.ones([numTrainingImages, 6]);
 // trainingData.tensor.output.dtype = "float32";