// --- defined by MNIST - do not change these ---------------------------------------
const PIXELS = 28; // images in data set are tiny
const PIXELSSQUARED = PIXELS * PIXELS;
var requiredPixels = 24;
// number of training and test exemplars in the data set:
const NOTRAIN = 60000;
const NOTEST = 10000;
//--- can modify all these --------------------------------------------------
// no of nodes in network
const noinput = PIXELSSQUARED;
const nohidden = 64;
const nooutput = 10;
const learningrate = 0.1; // default 0.1
// should we train every timestep or not
let do_training = true;
// how many to train and test per timestep
const TRAINPERSTEP = 30;
const TESTPERSTEP = 5;
// multiply it by this to magnify for display
const ZOOMFACTOR = 7;
const ZOOMPIXELS = ZOOMFACTOR * PIXELS;
// 3 rows of
// large image + 50 gap + small image
// 50 gap between rows
const canvaswidth = ( PIXELS + ZOOMPIXELS ) + 50;
const canvasheight = ( ZOOMPIXELS * 3 ) + 100;
const DOODLE_THICK = 18; // thickness of doodle lines
const DOODLE_BLUR = 3; // blur factor applied to doodles
let mnist;
// all data is loaded into this
// mnist.train_images
// mnist.train_labels
// mnist.test_images
// mnist.test_labels
let nn;
let trainrun = 1;
let train_index = 0;
let testrun = 1;
let test_index = 0;
let total_tests = 0;
let total_correct = 0;
// images in LHS:
let doodle, demo;
let doodle_exists = false;
let demo_exists = false;
let mousedrag = false; // are we in the middle of a mouse drag drawing?
// save inputs to global var to inspect
// type these names in console
var train_inputs, test_inputs, demo_inputs, doodle_inputs;
// Matrix.randomize() is changed to point to this. Must be defined by user of Matrix.
function randomWeight()
{
return ( AB.randomFloatAtoB ( -0.5, 0.5 ) );
// Coding Train default is -1 to 1
}
// make run header bigger
AB.headerCSS ( { "max-height": "95vh" } );
//--- start of AB.msgs structure: ---------------------------------------------------------
// We output a serious of AB.msgs to put data at various places in the run header
var thehtml;
// 1 Doodle header
thehtml = "<hr> <h1> 1. Doodle </h1> Top row: Doodle (left) and shrunk (right). <br> " +
" Draw your doodle number in top LHS. <button onclick='wipeDoodle();' class='normbutton' >Clear doodle</button> <br> ";
AB.msg ( thehtml, 1 );
// 2 Doodle variable data (guess)
// 3 Training header
thehtml = "<hr> <h1> 2. Training </h1> Middle row: Training image magnified (left) and original (right). <br> " +
" <button onclick='do_training = false;' class='normbutton' >Stop training</button> <br> ";
AB.msg ( thehtml, 3 );
// 4 variable training data
// 5 Testing header
thehtml = "<h3> Hidden tests </h3> " ;
AB.msg ( thehtml, 5 );
// 6 variable testing data
// 7 Demo header
thehtml = "<hr> <h1> 3. Demo </h1> Bottom row: Test image magnified (left) and original (right). <br>" +
" The network is <i>not</i> trained on any of these images. <br> " +
" <button onclick='makeDemo();' class='normbutton' >Demo test image</button> <br> ";
AB.msg ( thehtml, 7 );
// 8 Demo variable data (random demo ID)
// 9 Demo variable data (changing guess)
const greenspan = "<span style='font-weight:bold; font-size:x-large; color:darkgreen'> " ;
//--- end of AB.msgs structure: ---------------------------------------------------------
function imageToCenter(d, x) {
var path = oneDimensionTo2D(d, x);
var p = getIndexOfCorners(path, Number.MAX_VALUE, Number.MAX_VALUE, -1, -1);
/** @type {number} */
var aoff = Math.floor((x - p[3] - p[2]) / 2);
/** @type {number} */
var _open_dot = Math.floor((x - p[1] - p[0]) / 2);
/** @type {!Array} */
var a = [];
/** @type {number} */
var j = 0;
for (; j < x; j++) {
/** @type {!Array} */
a[j] = [];
/** @type {number} */
var k = 0;
for (; k < x; k++) {
/** @type {number} */
a[j][k] = 0;
}
}
j = p[2];
for (; j <= p[3]; j++) {
k = p[0];
for (; k <= p[1]; k++) {
a[j + aoff][k + _open_dot] = path[j][k];
}
}
return result = twoDto1D(a, x), result;
}
function twoDto1D(img, width) {
/** @type {!Array} */
var s_noiseLookup = [];
/** @type {number} */
var j = 0;
for (; j < width; j++) {
/** @type {number} */
var i = 0;
for (; i < width; i++) {
s_noiseLookup[j * width + i] = img[j][i];
}
}
return s_noiseLookup;
}
function getIndexOfCorners(val, min, max, n, result) {
/** @type {number} */
var k = 0;
for (; k < val.length; k++) {
var i = val[k].indexOf(255);
var value = val[k].lastIndexOf(255);
if (i >= 0 && i < max) {
max = i;
}
if (value >= 0 && value > result) {
result = value;
}
if (i >= 0 && k < min) {
/** @type {number} */
min = k;
}
if (i >= 0 && k > n) {
/** @type {number} */
n = k;
}
}
return [max, result, min, n];
}
function oneDimensionTo2D(length, scale) {
/** @type {!Array} */
var decTable = [];
/** @type {number} */
var i = 0;
for (; i < scale; i++) {
/** @type {!Array} */
decTable[i] = [];
/** @type {number} */
var s = 0;
for (; s < scale; s++) {
decTable[i][s] = length[4 * (i * scale + s)];
}
}
return decTable;
}
function guessTheDigit() {
var dst = doodle.get();
dst.resize(PIXELS, PIXELS);
dst.loadPixels();
let conf_shortcuts_icon = [];
for (let i = 0; i < PIXELSSQUARED; i++) {
/** @type {number} */
conf_shortcuts_icon[i] = dst.pixels[4 * i] / 255;
}
doodle_inputs = conf_shortcuts_icon;
var n = updateImageFormat(imageToCenter(dst.pixels, PIXELS), 24);
var o = find12(cnn.classifyImages([n])[0].values);
console.log ( cnn.classifyImages([n])[0] );
/** @type {string} */
thehtml = " I think it is: " + greenspan + o[0] + "</span> <br> No.2 guess is: " + greenspan + o[1] + "</span>";
AB.msg(thehtml, 2);
}
function loadNetworkFromJSON(networkJSON) {
cnn = new WebCNN;
if (void 0 != networkJSON.momentum) {
cnn.setMomentum(networkJSON.momentum);
}
if (void 0 != networkJSON.lambda) {
cnn.setLambda(networkJSON.lambda);
}
if (void 0 != networkJSON.learningRate) {
cnn.setLearningRate(networkJSON.learningRate);
}
/** @type {number} */
var layerIndex = 0;
for (; layerIndex < networkJSON.layers.length; ++layerIndex) {
let layerDesc = networkJSON.layers[layerIndex];
cnn.newLayer(layerDesc);
}
/** @type {number} */
layerIndex = 0;
for (; layerIndex < networkJSON.layers.length; ++layerIndex) {
let layerDesc = networkJSON.layers[layerIndex];
switch(networkJSON.layers[layerIndex].type) {
case LAYER_TYPE_CONV:
case LAYER_TYPE_FULLY_CONNECTED:
if (void 0 != layerDesc.weights && void 0 != layerDesc.biases) {
cnn.layers[layerIndex].setWeightsAndBiases(layerDesc.weights, layerDesc.biases);
}
}
}
return cnn.initialize(), cnn;
}
function setup() {
createCanvas(canvaswidth, canvasheight);
(doodle = createGraphics(ZOOMPIXELS, ZOOMPIXELS)).pixelDensity(1);
AB.loadingScreen();
$.getScript("/uploads/codingtrain/matrix.js", function() {
$.getScript("/uploads/codingtrain/nn.js", function() {
$.getScript("/uploads/codingtrain/mnist.js", function() {
$.getScript("uploads/danyal05/math.js", function() {
$.getScript("/uploads/danyal05/webcnn.js", function() {
$.getJSON("/uploads/danyal05/cnn_mnist_10_20_98accuracy.json", function(networkJSON, canCreateDiscussions) {
console.log("All JS loaded");
cnn = loadNetworkFromJSON(networkJSON);
(nn = new NeuralNetwork(noinput, nohidden, nooutput)).setLearningRate(learningrate);
loadData();
});
});
});
});
});
});
}
// load data set from local file (on this server)
function loadData()
{
loadMNIST ( function(data)
{
mnist = data;
console.log ("All data loaded into mnist object:")
console.log(mnist);
AB.removeLoading(); // if no loading screen exists, this does nothing
});
}
function updateImageFormat(password) {
return {
width : 24,
height : 24,
data : getRequiredImage(reduceImageSize(password)).pixels
};
}
function reduceImageSize(choices) {
/** @type {number} */
var caveWidth = PIXELS - 24;
/** @type {number} */
var offset = Math.floor(Math.random() * caveWidth);
/** @type {number} */
var repeaterItemIndex = Math.floor(Math.random() * caveWidth);
/** @type {number} */
var maxOffset = offset + 24;
/** @type {number} */
var i = repeaterItemIndex + 24;
/** @type {!Array} */
var returnChoices = [];
/** @type {number} */
var j = offset;
for (; j < maxOffset; j++) {
for (let k = repeaterItemIndex; k < i; k++) {
returnChoices.push(choices[j * PIXELS + k]);
}
}
return returnChoices;
}
function getRequiredImage(serverElements) {
let img = createImage(24, 24);
img.loadPixels();
for (let i = 0; i < 576; i++) {
let o = serverElements[i];
let index = 4 * i;
img.pixels[index + 0] = o;
img.pixels[index + 1] = o;
img.pixels[index + 2] = o;
/** @type {number} */
img.pixels[index + 3] = 255;
}
return img.updatePixels(), img;
}
function getImage ( img ) // make a P5 image object from a raw data array
{
let theimage = createImage (PIXELS, PIXELS); // make blank image, then populate it
theimage.loadPixels();
for (let i = 0; i < PIXELSSQUARED ; i++)
{
let bright = img[i];
let index = i * 4;
theimage.pixels[index + 0] = bright;
theimage.pixels[index + 1] = bright;
theimage.pixels[index + 2] = bright;
theimage.pixels[index + 3] = 255;
}
theimage.updatePixels();
return theimage;
}
function getInputs ( img ) // convert img array into normalised input array
{
let inputs = [];
for (let i = 0; i < PIXELSSQUARED ; i++)
{
let bright = img[i];
inputs[i] = bright / 255; // normalise to 0 to 1
}
return ( inputs );
}
function trainit(canCreateDiscussions) {
let id = mnist.train_images[train_index];
let PARAM_AUTOSTART = mnist.train_labels[train_index];
if (canCreateDiscussions) {
var img = getImage(id);
image(img, 0, ZOOMPIXELS + 50, ZOOMPIXELS, ZOOMPIXELS);
image(img, ZOOMPIXELS + 50, ZOOMPIXELS + 50, PIXELS, PIXELS);
}
let data = getInputs(id);
let parameters = [0, 0, 0, 0, 0, 0, 0, 0, 0, 0];
/** @type {number} */
parameters[PARAM_AUTOSTART] = 1;
train_inputs = data;
nn.train(data, parameters);
thehtml = " trainrun: " + trainrun + "<br> no: " + train_index;
AB.msg(thehtml, 4);
if (++train_index == NOTRAIN) {
/** @type {number} */
train_index = 0;
console.log("finished trainrun: " + trainrun);
trainrun++;
}
}
/**
* @return {undefined}
*/
function testItWithCNN() {
var test = mnist.test_images[test_index];
var e = mnist.test_labels[test_index];
var simpleResult = updateImageFormat(test, 24);
var inp = find12(cnn.classifyImages([simpleResult])[0].values);
total_tests++;
if (inp[0] == e) {
total_correct++;
}
/** @type {number} */
var e_total = total_correct / total_tests * 100;
/** @type {string} */
thehtml = " testrun: " + testrun + "<br> no: " + total_tests + " <br> correct: " + total_correct + "<br> score: " + greenspan + e_total.toFixed(2) + "</span>";
AB.msg(thehtml, 6);
if (++test_index == NOTEST) {
console.log("finished testrun: " + testrun + " score: " + e_total.toFixed(2));
testrun++;
/** @type {number} */
test_index = 0;
/** @type {number} */
total_tests = 0;
/** @type {number} */
total_correct = 0;
}
}
/**
* @return {undefined}
*/
function testit() {
let id = mnist.test_images[test_index];
let e = mnist.test_labels[test_index];
let sample = getInputs(id);
test_inputs = sample;
let place = findMax(nn.predict(sample));
total_tests++;
if (place == e) {
total_correct++;
}
let e_total = total_correct / total_tests * 100;
/** @type {string} */
thehtml = " testrun: " + testrun + "<br> no: " + total_tests + " <br> correct: " + total_correct + "<br> score: " + greenspan + e_total.toFixed(2) + "</span>";
AB.msg(thehtml, 6);
if (++test_index == NOTEST) {
console.log("finished testrun: " + testrun + " score: " + e_total.toFixed(2));
testrun++;
/** @type {number} */
test_index = 0;
/** @type {number} */
total_tests = 0;
/** @type {number} */
total_correct = 0;
}
}
//--- find no.1 (and maybe no.2) output nodes ---------------------------------------
// (restriction) assumes array values start at 0 (which is true for output nodes)
function find12 (a) // return array showing indexes of no.1 and no.2 values in array
{
let no1 = 0;
let no2 = 0;
let no1value = 0;
let no2value = 0;
for (let i = 0; i < a.length; i++)
{
if (a[i] > no1value) // new no1
{
// old no1 becomes no2
no2 = no1;
no2value = no1value;
// now put in the new no1
no1 = i;
no1value = a[i];
}
else if (a[i] > no2value) // new no2
{
no2 = i;
no2value = a[i];
}
}
var b = [ no1, no2 ];
return b;
}
// just get the maximum - separate function for speed - done many times
// find our guess - the max of the output nodes array
function findMax (a)
{
let no1 = 0;
let no1value = 0;
for (let i = 0; i < a.length; i++)
{
if (a[i] > no1value)
{
no1 = i;
no1value = a[i];
}
}
return no1;
}
// --- the draw function -------------------------------------------------------------
// every step:
function draw() {
if (void 0 !== mnist) {
if (background("black"), do_training) {
for (let t = 0; t < TRAINPERSTEP; t++) {
trainit(0 == t);
}
for (let t = 0; t < TESTPERSTEP; t++) {
testItWithCNN();
}
}
if (demo_exists && (drawDemo(), guessDemo()), doodle_exists && (drawDoodle(), guessTheDigit()), mouseIsPressed) {
var left = ZOOMPIXELS + 20;
if (mouseX < left && mouseY < left && pmouseX < left && pmouseY < left) {
/** @type {boolean} */
mousedrag = true;
/** @type {boolean} */
doodle_exists = true;
doodle.stroke("white");
doodle.strokeWeight(DOODLE_THICK);
doodle.line(mouseX, mouseY, pmouseX, pmouseY);
}
} else {
if (mousedrag) {
/** @type {boolean} */
mousedrag = false;
doodle.filter(BLUR, DOODLE_BLUR);
}
}
}
}
//--- demo -------------------------------------------------------------
// demo some test image and predict it
// get it from test set so have not used it in training
function makeDemo()
{
demo_exists = true;
var i = AB.randomIntAtoB ( 0, NOTEST - 1 );
demo = mnist.test_images[i];
var label = mnist.test_labels[i];
thehtml = "Test image no: " + i + "<br>" +
"Classification: " + label + "<br>" ;
AB.msg ( thehtml, 8 );
// type "demo" in console to see raw data
}
function drawDemo()
{
var theimage = getImage ( demo );
// console.log (theimage);
image ( theimage, 0, canvasheight - ZOOMPIXELS, ZOOMPIXELS, ZOOMPIXELS ); // magnified
image ( theimage, ZOOMPIXELS+50, canvasheight - ZOOMPIXELS, PIXELS, PIXELS ); // original
}
function guessDemo()
{
let inputs = getInputs ( demo );
demo_inputs = inputs; // can inspect in console
let prediction = nn.predict(inputs); // array of outputs
let guess = findMax(prediction); // the top output
thehtml = " I think it is: " + greenspan + guess + "</span>" ;
AB.msg ( thehtml, 9 );
}
//--- doodle -------------------------------------------------------------
function drawDoodle()
{
// doodle is createGraphics not createImage
let theimage = doodle.get();
// console.log (theimage);
image ( theimage, 0, 0, ZOOMPIXELS, ZOOMPIXELS ); // original
image ( theimage, ZOOMPIXELS+50, 0, PIXELS, PIXELS ); // shrunk
}
function guessDoodle()
{
// doodle is createGraphics not createImage
let img = doodle.get();
img.resize ( PIXELS, PIXELS );
img.loadPixels();
// set up inputs
let inputs = [];
for (let i = 0; i < PIXELSSQUARED ; i++)
{
inputs[i] = img.pixels[i * 4] / 255;
}
doodle_inputs = inputs; // can inspect in console
// feed forward to make prediction
let prediction = nn.predict(inputs); // array of outputs
let b = find12(prediction); // get no.1 and no.2 guesses
thehtml = " I think it is: " + greenspan + b[0] + "</span> <br>" +
" Maybe ...: " + greenspan + b[1] + "</span>";
AB.msg ( thehtml, 2 );
}
function wipeDoodle()
{
doodle_exists = false;
doodle.background('black');
}
// --- debugging --------------------------------------------------
// in console
// showInputs(demo_inputs);
// showInputs(doodle_inputs);
function showInputs ( inputs )
// display inputs row by row, corresponding to square of pixels
{
var str = "";
for (let i = 0; i < inputs.length; i++)
{
if ( i % PIXELS == 0 ) str = str + "\n"; // new line for each row of pixels
var value = inputs[i];
str = str + " " + value.toFixed(2) ;
}
console.log (str);
}