// Cloned by Tony Forde on 15 Nov 2021 from World "XOR multi-layer network" by "Coding Train" project 
// Please leave this clone trail here.
 

// XOR multi-layer network

// Port from:
// https://github.com/CodingTrain/Toy-Neural-Network-JS/tree/master/examples/xor
// with modifications 

// libraries from:
// https://github.com/CodingTrain/Toy-Neural-Network-JS/tree/master/lib
// ported to here:
// https://ancientbrain.com/uploads.php?userid=codingtrain 



//=== Tweaker's box ============================================
// TEST: Change the number of nodes in each layer. Defaults:
// const noinput = 2;
// const nohidden = 6;
// const nooutput = 1;
// number of nodes in each layer:
const noinput = 2; // Can only be 2 as inputs are 1, 0
const nohidden = 6;
const nooutput = 1; // Can only be one output: 1 or 0

// TEST: Change the training data to represent something other than XOR. Defaults:
// let training_data = [
//   {    inputs: [0, 0],    outputs: [0]  },
//   {    inputs: [0, 1],    outputs: [1]  },
//   {    inputs: [1, 0],    outputs: [1]  },
//   {    inputs: [1, 1],    outputs: [0]  }
// ];
// define the exemplars to learn from:
let training_data = [
  {    inputs: [0, 0],    outputs: [0]  },
  {    inputs: [0, 1],    outputs: [1]  },
  {    inputs: [1, 0],    outputs: [1]  },
  {    inputs: [1, 1],    outputs: [0]  }
];

var nn;     // global var 

// TEST: Change the learning rate (default 0.2)
// How much should the network change every time we get something wrong?
const LEARNING_RATE_MAX = 0.4; 
const LEARNING_RATE_FLOOR = 0.15; 
const LEARNING_RATE_REDUCTION = 0.99553; 
var learningrate = LEARNING_RATE_MAX; 

// train this number of times per draw()
const notrain = 10;

// Take screenshot on this step:
AB.screenshotStep  = 200;   


// divide 0,1 into squares 
// show all squares or just the corner squares:
var showall = true;

const canvassize    = 400;
const squaresize    = 40;

const cols          = 10 ;
const rows          = 10;


// Matrix.randomize() is changed to point to this. Must be defined by user of Matrix. 

// TEST: Change the lower and upper weight range, set it to all 0, set it to a constant...

function randomWeight()
{
    // Default:
    // return ( AB.randomFloatAtoB ( -0.5, 0.5 ) );
    // Coding Train default is -1 to 1
    return ( AB.randomFloatAtoB ( -0.5, 0.5 ) );
}    

//=== End of tweaker's box ============================================




function setup() 
{
  createCanvas (canvassize, canvassize);
  
   $.getScript ( "/uploads/codingtrain/matrix.js", function()
   {
        $.getScript ( "/uploads/codingtrain/nn.js", function()
        {
            nn = new NeuralNetwork ( noinput, nohidden, nooutput );
        });
   });
}



function draw() 
{
  // check if libraries loaded yet:
  if ( typeof nn == 'undefined' ) return;
  
  learningrate = learningrate * LEARNING_RATE_REDUCTION;
  if (learningrate < LEARNING_RATE_REDUCTION)
    learningrate = LEARNING_RATE_REDUCTION;
    console.log('learning rate = ' + learningrate);
  nn.setLearningRate ( learningrate );

  background ('#ffffcc'); 

  // train n times 
  for (let i = 0; i < notrain ; i++) 
  {
    let data = random ( training_data );
    nn.train ( data.inputs, data.outputs );
  }

// draw either some squares or all squares:

  if ( showall )
  {
    // redraw all squares each time round
    for (let i = 0; i < cols; i++) 
        for (let j = 0; j < rows; j++) 
            drawquare ( i, j );
  }
  else
  {
    // redraw just the 4 squares 
    for ( let i = 0; i < cols; i = i + cols-1 ) 
        for ( let j = 0; j < rows; j = j + rows-1 ) 
            drawquare ( i, j );
  }
}  
     
 
function drawquare ( i, j )
{
      let x1 = i / cols;
      let x2 = j / rows;
      let inputs = [x1, x2];
      let y = nn.predict(inputs);
      // console.log ( "input (" +x1 + "," + x2 + ") output " + y );
      
      strokeWeight(2);
      stroke('black');
      fill ( y * 255 );      // 0 is black, 1 is white 
      
       rect ( i * squaresize, j * squaresize, squaresize, squaresize );
}