w2v-server: move js to folder
diff --git a/js/som.js b/js/som.js
new file mode 100644
index 0000000..ca16d28
--- /dev/null
+++ b/js/som.js
@@ -0,0 +1,286 @@
+// Javascript implementation pf Kohonen's Self Organizing Map
+// Based on http://www.ai-junkie.com/ann/som/som1.html
+
+var mapWidth = 800,
+ mapHeight = 800;
+
+function getDistance(weight, inputVector) {
+ var distance = 0;
+ for (var i = 0; i <weight.length; i++) {
+ distance += (inputVector[i] - weight[i]) * (inputVector[i] - weight[i]);
+ }
+ return Math.sqrt(distance);
+}
+
+function makeRandomWeights(vSize, eSize) {
+ var weights = [];
+ if(typeof ArrayBuffer === 'undefined') {
+ // lacking browser support
+ while (weights.length < vSize) {
+ var arr = new Array(eSize);
+ for(var i = 0; i < eSize; i++) { arr[i]= Math.random(); }
+ weights.push(arr);
+ }
+ } else {
+ while (weights.length < vSize) {
+ var arr = new Float64Array(eSize);
+ for(var i = 0; i < eSize; i++) { arr[i]= Math.random(); }
+ weights.push(arr);
+ }
+ }
+ return weights;
+}
+
+function getBMUIndex(weights, target) {
+ var BMUIndex = 0;
+ var bestScore = 99999;
+
+ for (i=0; i < weights.length; i++) {
+ distance = getDistance(weights[i], target);
+ if (distance < bestScore) {
+ bestScore = distance;
+ BMUIndex = i;
+ }
+ }
+ return BMUIndex;
+}
+
+function convertIndexToXY(idx, dimW) {
+ var x = parseInt(idx % dimW,10);
+ var y = parseInt((idx / dimW),10);
+ return [x,y];
+}
+
+
+function getEucledianDistance(coord1, coord2) {
+ return (coord1[0] - coord2[0]) * (coord1[0] - coord2[0]) + (coord1[1] - coord2[1]) * (coord1[1] - coord2[1]);
+}
+
+// utilitity that creates contiguous vector of zeros of size n
+ var zeros = function(n) {
+ if(typeof(n)==='undefined' || isNaN(n)) { return []; }
+ if(typeof ArrayBuffer === 'undefined') {
+ // lacking browser support
+ var arr = new Array(n);
+ for(var i=0;i<n;i++) { arr[i]= 0; }
+ return arr;
+ } else {
+ return new Float64Array(n); // typed arrays are faster
+ }
+ }
+
+// compute L2 distance between two vectors
+var L2 = function(x1, x2) {
+ var D = x1.length;
+ var d = 0;
+ for(var i=0;i<D;i++) {
+ var x1i = x1[i];
+ var x2i = x2[i];
+ d += (x1i-x2i)*(x1i-x2i);
+ }
+ return d;
+}
+
+// compute pairwise distance in all vectors in X
+var xtod = function(X) {
+ var N = X.length;
+ var dist = zeros(N * N); // allocate contiguous array
+ for(var i=0;i<N;i++) {
+ for(var j=i+1;j<N;j++) {
+ var d = L2(X[i], X[j]);
+ dist[i*N+j] = d;
+ dist[j*N+i] = d;
+ }
+ }
+ return dist;
+}
+
+function dotproduct(a,b) {
+ var n = 0, lim = Math.min(a.length,b.length);
+ for (var i = 0; i < lim; i++) n += a[i] * b[i];
+ return n;
+ }
+
+function vecsum(a,b) {
+ var lim = a.length;
+ var sum = new Array(lim);
+ for (var i = 0; i < lim; i++) sum[i] = a[i] + b[i];
+ return sum;
+ }
+
+function norm2(a) {var sumsqr = 0; for (var i = 0; i < a.length; i++) sumsqr += a[i]*a[i]; return Math.sqrt(sumsqr);}
+
+function cosine_sim(x, y) {
+ xnorm = norm2(x);
+ if(!xnorm) return 0;
+ ynorm = norm2(y);
+ if(!ynorm) return 0;
+ return dotproduct(x, y) / (xnorm * ynorm);
+}
+
+function makeSOM(data, training_iterations) {
+ var dimW = 6;
+ var dimH = 6;
+
+ var radius = (dimW * dimH) / 2;
+ var learning_rate = 1;
+ var time_constant = training_iterations / Math.log(radius);
+ var inputs = xtod(data.vecs);
+ var dimI = data.vecs.length;
+ var weights = makeRandomWeights(dimW * dimH, dimI);
+ var radius_decaying = 0;
+ var learning_rate_decaying = 0;
+ var svg;
+ var no_targets = (data.target.match(/.[ |]+./g) || []).length+1;
+// var avg, avgsim1, avgsim2, minsim;
+ var refIndex;
+ var colorScale;
+
+ if(no_targets > 1) {
+ refIndex=1;
+ colorScale = d3.scale.linear()
+ .range(['green', 'yellow', 'red']) // or use hex values
+ .domain([-1, 0, 1]);
+
+ // avg = vecsum(inputs.slice(0, dimI), inputs.slice(dimI, 2*dimI));
+ // avgsim1 = cosine_sim(inputs.slice(0, dimI), avg);
+ // avgsim2 = cosine_sim(inputs.slice(dimI, 2*dimI), avg);
+
+ $("#somcolor2").css("background-color", colorScale(0));
+ $("#somcolor1").css("background-color", colorScale(-1));
+ $("#somcolor3").css("background-color", colorScale(1));
+ } else {
+ refIndex = data.words.length-1;
+ colorScale = d3.scale.linear()
+ .range(['white', 'red'])
+ .domain([-1, 1]);
+ $("#somcolor1").css("background-color", colorScale(1));
+ $("#somcolor3").css("background-color", colorScale(-1));
+ }
+
+ $("#somword1").html(data.words[0]);
+ $("#somword2").html(data.words[refIndex]);
+ minsim = cosine_sim(inputs.slice(0, dimI), inputs.slice(refIndex*dimI, (refIndex+1)*dimI));
+
+ var itdom = document.getElementById("iterations");
+
+ var div = d3.select("#som2");
+
+ data.coords = [];
+ for(var i=0; i< data.words.length; i++) {
+ data.coords[i] = [Math.floor(dimW/2), Math.floor(dimH/2)];
+ }
+
+ svg = div.append("svg")
+ .attr("width", mapWidth)
+ .attr("height", mapHeight);
+
+ var rects = svg.selectAll(".r")
+ .data(weights)
+ .enter().append("rect")
+ .attr("class", "r")
+ .attr("width", mapWidth/dimW)
+ .attr("height", mapHeight/dimH)
+ .attr("fill", "white")
+ .attr("z-index", "-1")
+ .attr("x", function(d, i) { return (i % dimW) * (mapWidth/dimW);})
+ .attr("y", function(d, i) { return (Math.floor(i / dimW) * (mapWidth/dimW)); })
+
+
+ var g = svg.selectAll(".b")
+ .data(data.words)
+ .enter().append("g")
+ .attr("class", "u");
+ g.append("a")
+ .attr("xlink:href", function(word) {return data.urlprefix+word;})
+ .attr("title", function(d, i) {
+ return "rank: "+i +" "+"freq. rank: "+data.ranks[i].toString().replace(/\B(?=(\d{3})+(?!\d))/g, ",");
+ })
+ .append("text")
+ .attr("text-anchor", "bottom")
+ .attr("font-size", 12)
+ .attr("fill", function(d) {
+ if(data.target.indexOf(" "+d+" ") >= 0) {
+ return "blue";
+ } else {
+ return "#333"
+ }
+ })
+ .text(function(d) { return d; });
+
+ var som_interval = setInterval(somStep, 0);
+ var it=0;
+
+ function updateSOM() {
+ var oc = [];
+ for(var x = 0; x < dimW; x++) {
+ for(var y = 0; y < dimH; y++) {
+ oc[y*dimW+x]=1;
+ }
+ }
+ svg.selectAll('.u')
+ .data(data.coords)
+ .transition()
+ .attr("transform", function(d, i) {
+ return "translate(" +
+ (d[0]*(mapWidth/dimW)+4) + "," +
+ (d[1]*(mapHeight/dimH)+oc[d[1]*dimW+d[0]]++*14+4) + ")"; });
+
+ var colorFun = function(d, i) {
+ var sim1=cosine_sim(d, inputs.slice(0, dimI));
+ var sim2=cosine_sim(d, inputs.slice(dimI, 2*dimI));
+ var col;
+// col = (sim1-avgsim1)/(1-avgsim1)-(sim2-avgsim2)/(1-avgsim2);
+ col = (sim2-sim1)/(1-minsim);
+// console.log(Math.floor(i/dimW)+","+i%dimW+":"+(sim1-minsim)/(1-minsim)+ " " + (sim2-minsim)/(1-minsim) + "--> "+ col);
+ if(col > 1) col=1;
+ if(col < -1) col=-1;
+ return colorScale(col);
+ };
+
+ if(it>training_iterations*.6) {
+ svg.selectAll(".r")
+ .data(weights)
+ .transition()
+ .attr("fill", colorFun);
+ }
+ }
+
+ function somStep() {
+ if(it++ >= training_iterations) {
+ updateSOM();
+ clearInterval(som_interval);
+ return;
+ }
+ itdom.innerHTML = it;
+ radius_decaying = radius * Math.exp(-it/time_constant);
+ learning_rate_decaying = learning_rate * Math.exp(-it/time_constant);
+ //learning_rate_decaying = learning_rate * Math.exp(-it/training_iterations);
+
+ //pick a random input to train
+ var current=Math.floor(Math.random()*dimI)
+ var iv = inputs.slice(current*dimI, (current+1)*dimI);
+ // Determine the BMU
+ BMUIdx = getBMUIndex(weights, iv);
+ var coord1 = convertIndexToXY(BMUIdx, dimW);
+ data.coords[current] = coord1;
+ var widthSq = radius_decaying * radius_decaying;
+ for (var v in weights) {
+ var coord2 = convertIndexToXY(v, dimW);
+ var dist = getEucledianDistance(coord1, coord2);
+ // Determine if the weight is within the training radius
+ if (dist < widthSq) {
+ // console.log(dist, learning_rate_decaying, radius_decaying, it);
+ influence = Math.exp(-dist/(2*widthSq));
+ for (vidx = 0;vidx<weights[v].length;vidx++) {
+ weights[v][vidx] += influence * learning_rate_decaying * (iv[vidx] - weights[v][vidx]);
+ }
+ }
+ }
+// }
+ if(it % 10 == 0) {
+ updateSOM();
+ }
+ }
+
+}
diff --git a/js/tsne.js b/js/tsne.js
new file mode 100644
index 0000000..57b3878
--- /dev/null
+++ b/js/tsne.js
@@ -0,0 +1,371 @@
+// create main global object
+var tsnejs = tsnejs || { REVISION: 'ALPHA' };
+
+(function(global) {
+ "use strict";
+
+ // utility function
+ var assert = function(condition, message) {
+ if (!condition) { throw message || "Assertion failed"; }
+ }
+
+ // syntax sugar
+ var getopt = function(opt, field, defaultval) {
+ if(opt.hasOwnProperty(field)) {
+ return opt[field];
+ } else {
+ return defaultval;
+ }
+ }
+
+ // return 0 mean unit standard deviation random number
+ var return_v = false;
+ var v_val = 0.0;
+ var gaussRandom = function() {
+ if(return_v) {
+ return_v = false;
+ return v_val;
+ }
+ var u = 2*Math.random()-1;
+ var v = 2*Math.random()-1;
+ var r = u*u + v*v;
+ if(r == 0 || r > 1) return gaussRandom();
+ var c = Math.sqrt(-2*Math.log(r)/r);
+ v_val = v*c; // cache this for next function call for efficiency
+ return_v = true;
+ return u*c;
+ }
+
+ // return random normal number
+ var randn = function(mu, std){ return mu+gaussRandom()*std; }
+
+ // utilitity that creates contiguous vector of zeros of size n
+ var zeros = function(n) {
+ if(typeof(n)==='undefined' || isNaN(n)) { return []; }
+ if(typeof ArrayBuffer === 'undefined') {
+ // lacking browser support
+ var arr = new Array(n);
+ for(var i=0;i<n;i++) { arr[i]= 0; }
+ return arr;
+ } else {
+ return new Float64Array(n); // typed arrays are faster
+ }
+ }
+
+ // utility that returns 2d array filled with random numbers
+ // or with value s, if provided
+ var randn2d = function(n,d,s) {
+ var uses = typeof s !== 'undefined';
+ var x = [];
+ for(var i=0;i<n;i++) {
+ var xhere = [];
+ for(var j=0;j<d;j++) {
+ if(uses) {
+ xhere.push(s);
+ } else {
+ xhere.push(randn(0.0, 1e-4));
+ }
+ }
+ x.push(xhere);
+ }
+ return x;
+ }
+
+ // compute L2 distance between two vectors
+ var L2 = function(x1, x2) {
+ var D = x1.length;
+ var d = 0;
+ for(var i=0;i<D;i++) {
+ var x1i = x1[i];
+ var x2i = x2[i];
+ d += (x1i-x2i)*(x1i-x2i);
+ }
+ return d;
+ }
+
+ // compute pairwise distance in all vectors in X
+ var xtod = function(X) {
+ var N = X.length;
+ var dist = zeros(N * N); // allocate contiguous array
+ for(var i=0;i<N;i++) {
+ for(var j=i+1;j<N;j++) {
+ var d = L2(X[i], X[j]);
+ dist[i*N+j] = d;
+ dist[j*N+i] = d;
+ }
+ }
+ return dist;
+ }
+
+ // compute (p_{i|j} + p_{j|i})/(2n)
+ var d2p = function(D, perplexity, tol) {
+ var Nf = Math.sqrt(D.length); // this better be an integer
+ var N = Math.floor(Nf);
+ assert(N === Nf, "D should have square number of elements.");
+ var Htarget = Math.log(perplexity); // target entropy of distribution
+ var P = zeros(N * N); // temporary probability matrix
+
+ var prow = zeros(N); // a temporary storage compartment
+ for(var i=0;i<N;i++) {
+ var betamin = -Infinity;
+ var betamax = Infinity;
+ var beta = 1; // initial value of precision
+ var done = false;
+ var maxtries = 50;
+
+ // perform binary search to find a suitable precision beta
+ // so that the entropy of the distribution is appropriate
+ var num = 0;
+ while(!done) {
+ //debugger;
+
+ // compute entropy and kernel row with beta precision
+ var psum = 0.0;
+ for(var j=0;j<N;j++) {
+ var pj = Math.exp(- D[i*N+j] * beta);
+ if(i===j) { pj = 0; } // we dont care about diagonals
+ prow[j] = pj;
+ psum += pj;
+ }
+ // normalize p and compute entropy
+ var Hhere = 0.0;
+ for(var j=0;j<N;j++) {
+ var pj = prow[j] / psum;
+ prow[j] = pj;
+ if(pj > 1e-7) Hhere -= pj * Math.log(pj);
+ }
+
+ // adjust beta based on result
+ if(Hhere > Htarget) {
+ // entropy was too high (distribution too diffuse)
+ // so we need to increase the precision for more peaky distribution
+ betamin = beta; // move up the bounds
+ if(betamax === Infinity) { beta = beta * 2; }
+ else { beta = (beta + betamax) / 2; }
+
+ } else {
+ // converse case. make distrubtion less peaky
+ betamax = beta;
+ if(betamin === -Infinity) { beta = beta / 2; }
+ else { beta = (beta + betamin) / 2; }
+ }
+
+ // stopping conditions: too many tries or got a good precision
+ num++;
+ if(Math.abs(Hhere - Htarget) < tol) { done = true; }
+ if(num >= maxtries) { done = true; }
+ }
+
+ // console.log('data point ' + i + ' gets precision ' + beta + ' after ' + num + ' binary search steps.');
+ // copy over the final prow to P at row i
+ for(var j=0;j<N;j++) { P[i*N+j] = prow[j]; }
+
+ } // end loop over examples i
+
+ // symmetrize P and normalize it to sum to 1 over all ij
+ var Pout = zeros(N * N);
+ var N2 = N*2;
+ for(var i=0;i<N;i++) {
+ for(var j=0;j<N;j++) {
+ Pout[i*N+j] = Math.max((P[i*N+j] + P[j*N+i])/N2, 1e-100);
+ }
+ }
+
+ return Pout;
+ }
+
+ // helper function
+ function sign(x) { return x > 0 ? 1 : x < 0 ? -1 : 0; }
+
+ var tSNE = function(opt) {
+ var opt = opt || {};
+ this.perplexity = getopt(opt, "perplexity", 30); // effective number of nearest neighbors
+ this.dim = getopt(opt, "dim", 2); // by default 2-D tSNE
+ this.epsilon = getopt(opt, "epsilon", 10); // learning rate
+
+ this.iter = 0;
+ }
+
+ tSNE.prototype = {
+
+ // this function takes a set of high-dimensional points
+ // and creates matrix P from them using gaussian kernel
+ initDataRaw: function(X) {
+ var N = X.length;
+ var D = X[0].length;
+ assert(N > 0, " X is empty? You must have some data!");
+ assert(D > 0, " X[0] is empty? Where is the data?");
+ var dists = xtod(X); // convert X to distances using gaussian kernel
+ this.P = d2p(dists, this.perplexity, 1e-4); // attach to object
+ this.N = N; // back up the size of the dataset
+ this.initSolution(); // refresh this
+ },
+
+ // this function takes a given distance matrix and creates
+ // matrix P from them.
+ // D is assumed to be provided as a list of lists, and should be symmetric
+ initDataDist: function(D) {
+ var N = D.length;
+ assert(N > 0, " X is empty? You must have some data!");
+ // convert D to a (fast) typed array version
+ var dists = zeros(N * N); // allocate contiguous array
+ for(var i=0;i<N;i++) {
+ for(var j=i+1;j<N;j++) {
+ var d = D[i][j];
+ dists[i*N+j] = d;
+ dists[j*N+i] = d;
+ }
+ }
+ this.P = d2p(dists, this.perplexity, 1e-4);
+ this.N = N;
+ this.initSolution(); // refresh this
+ },
+
+ // (re)initializes the solution to random
+ initSolution: function() {
+ // generate random solution to t-SNE
+ this.Y = randn2d(this.N, this.dim); // the solution
+ this.gains = randn2d(this.N, this.dim, 1.0); // step gains to accelerate progress in unchanging directions
+ this.ystep = randn2d(this.N, this.dim, 0.0); // momentum accumulator
+ this.iter = 0;
+ },
+
+ // return pointer to current solution
+ getSolution: function() {
+ return this.Y;
+ },
+
+ // perform a single step of optimization to improve the embedding
+ step: function() {
+ this.iter += 1;
+ var N = this.N;
+
+ var cg = this.costGrad(this.Y); // evaluate gradient
+ var cost = cg.cost;
+ var grad = cg.grad;
+
+ // perform gradient step
+ var ymean = zeros(this.dim);
+ for(var i=0;i<N;i++) {
+ for(var d=0;d<this.dim;d++) {
+ var gid = grad[i][d];
+ var sid = this.ystep[i][d];
+ var gainid = this.gains[i][d];
+
+ // compute gain update
+ var newgain = sign(gid) === sign(sid) ? gainid * 0.8 : gainid + 0.2;
+ if(gainid < 0.01) gainid = 0.01; // clamp
+ this.gains[i][d] = newgain; // store for next turn
+
+ // compute momentum step direction
+ var momval = this.iter < 250 ? 0.5 : 0.8;
+ var newsid = momval * sid - this.epsilon * newgain * grad[i][d];
+ this.ystep[i][d] = newsid; // remember the step we took
+
+ // step!
+ this.Y[i][d] += newsid;
+
+ ymean[d] += this.Y[i][d]; // accumulate mean so that we can center later
+ }
+ }
+
+ // reproject Y to be zero mean
+ for(var i=0;i<N;i++) {
+ for(var d=0;d<this.dim;d++) {
+ this.Y[i][d] -= ymean[d]/N;
+ }
+ }
+
+ //if(this.iter%100===0) console.log('iter ' + this.iter + ', cost: ' + cost);
+ return cost; // return current cost
+ },
+
+ // for debugging: gradient check
+ debugGrad: function() {
+ var N = this.N;
+
+ var cg = this.costGrad(this.Y); // evaluate gradient
+ var cost = cg.cost;
+ var grad = cg.grad;
+
+ var e = 1e-5;
+ for(var i=0;i<N;i++) {
+ for(var d=0;d<this.dim;d++) {
+ var yold = this.Y[i][d];
+
+ this.Y[i][d] = yold + e;
+ var cg0 = this.costGrad(this.Y);
+
+ this.Y[i][d] = yold - e;
+ var cg1 = this.costGrad(this.Y);
+
+ var analytic = grad[i][d];
+ var numerical = (cg0.cost - cg1.cost) / ( 2 * e );
+ console.log(i + ',' + d + ': gradcheck analytic: ' + analytic + ' vs. numerical: ' + numerical);
+
+ this.Y[i][d] = yold;
+ }
+ }
+ },
+
+ // return cost and gradient, given an arrangement
+ costGrad: function(Y) {
+ var N = this.N;
+ var dim = this.dim; // dim of output space
+ var P = this.P;
+
+ var pmul = this.iter < 100 ? 4 : 1; // trick that helps with local optima
+
+ // compute current Q distribution, unnormalized first
+ var Qu = zeros(N * N);
+ var qsum = 0.0;
+ for(var i=0;i<N;i++) {
+ for(var j=i+1;j<N;j++) {
+ var dsum = 0.0;
+ for(var d=0;d<dim;d++) {
+ var dhere = Y[i][d] - Y[j][d];
+ dsum += dhere * dhere;
+ }
+ var qu = 1.0 / (1.0 + dsum); // Student t-distribution
+ Qu[i*N+j] = qu;
+ Qu[j*N+i] = qu;
+ qsum += 2 * qu;
+ }
+ }
+ // normalize Q distribution to sum to 1
+ var NN = N*N;
+ var Q = zeros(NN);
+ for(var q=0;q<NN;q++) { Q[q] = Math.max(Qu[q] / qsum, 1e-100); }
+
+ var cost = 0.0;
+ var grad = [];
+ for(var i=0;i<N;i++) {
+ var gsum = new Array(dim); // init grad for point i
+ for(var d=0;d<dim;d++) { gsum[d] = 0.0; }
+ for(var j=0;j<N;j++) {
+ cost += - P[i*N+j] * Math.log(Q[i*N+j]); // accumulate cost (the non-constant portion at least...)
+ var premult = 4 * (pmul * P[i*N+j] - Q[i*N+j]) * Qu[i*N+j];
+ for(var d=0;d<dim;d++) {
+ gsum[d] += premult * (Y[i][d] - Y[j][d]);
+ }
+ }
+ grad.push(gsum);
+ }
+
+ return {cost: cost, grad: grad};
+ }
+ }
+
+ global.tSNE = tSNE; // export tSNE class
+})(tsnejs);
+
+
+// export the library to window, or to module in nodejs
+(function(lib) {
+ "use strict";
+ if (typeof module === "undefined" || typeof module.exports === "undefined") {
+ window.tsnejs = lib; // in ordinary browser attach library to window
+ } else {
+ module.exports = lib; // in nodejs
+ }
+})(tsnejs);