var SVC = function(nClasses, nRows, vectors, coefficients, intercepts, weights, kernel, gamma, coef0, degree) {

    this.nClasses = nClasses;
    this.classes = new Array(nClasses);
    for (var i = 0; i < nClasses; i++) {
        this.classes[i] = i;
    }
    this.nRows = nRows;
    this.vectors = vectors;
    this.coefficients = coefficients;
    this.intercepts = intercepts;
    this.weights = weights;
    this.kernel = kernel.toUpperCase();
    this.gamma = gamma;
    this.coef0 = coef0;
    this.degree = degree;

    this.predict = function(features) {
    
        var kernels = new Array(vectors.length);
        var kernel;
        switch (this.kernel) {
            case 'LINEAR':
                // <x,x'>
                for (var i = 0; i < this.vectors.length; i++) {
                    kernel = 0.;
                    for (var j = 0; j < this.vectors[i].length; j++) {
                        kernel += this.vectors[i][j] * features[j];
                    }
                    kernels[i] = kernel;
                }
                break;
            case 'POLY':
                // (y<x,x'>+r)^d
                for (var i = 0; i < this.vectors.length; i++) {
                    kernel = 0.;
                    for (var j = 0; j < this.vectors[i].length; j++) {
                        kernel += this.vectors[i][j] * features[j];
                    }
                    kernels[i] = Math.pow((this.gamma * kernel) + this.coef0, this.degree);
                }
                break;
            case 'RBF':
                // exp(-y|x-x'|^2)
                for (var i = 0; i < this.vectors.length; i++) {
                    kernel = 0.;
                    for (var j = 0; j < this.vectors[i].length; j++) {
                        kernel += Math.pow(this.vectors[i][j] - features[j], 2);
                    }
                    kernels[i] = Math.exp(-this.gamma * kernel);
                }
                break;
            case 'SIGMOID':
                // tanh(y<x,x'>+r)
                for (var i = 0; i < this.vectors.length; i++) {
                    kernel = 0.;
                    for (var j = 0; j < this.vectors[i].length; j++) {
                        kernel += this.vectors[i][j] * features[j];
                    }
                    kernels[i] = Math.tanh((this.gamma * kernel) + this.coef0);
                }
                break;
        }
    
        var starts = new Array(this.nRows);
        for (var i = 0; i < this.nRows; i++) {
            if (i != 0) {
                var start = 0;
                for (var j = 0; j < i; j++) {
                    start += this.weights[j];
                }
                starts[i] = start;
            } else {
                starts[0] = 0;
            }
        }
    
        var ends = new Array(this.nRows);
        for (var i = 0; i < this.nRows; i++) {
            ends[i] = this.weights[i] + starts[i];
        }
    
        if (this.nClasses == 2) {
    
            for (var i = 0; i < kernels.length; i++) {
                kernels[i] = -kernels[i];
            }
    
            var decision = 0.;
            for (var k = starts[1]; k < ends[1]; k++) {
                decision += kernels[k] * this.coefficients[0][k];
            }
            for (var k = starts[0]; k < ends[0]; k++) {
                decision += kernels[k] * this.coefficients[0][k];
            }
            decision += this.intercepts[0];
    
            if (decision > 0) {
                return 0;
            }
            return 1;
    
        }
    
        var decisions = new Array(this.intercepts.length);
        for (var i = 0, d = 0, l = this.nRows; i < l; i++) {
            for (var j = i + 1; j < l; j++) {
                var tmp = 0.;
                for (var k = starts[j]; k < ends[j]; k++) {
                    tmp += this.coefficients[i][k] * kernels[k];
                }
                for (var k = starts[i]; k < ends[i]; k++) {
                    tmp += this.coefficients[j - 1][k] * kernels[k];
                }
                decisions[d] = tmp + this.intercepts[d];
                d++;
            }
        }
    
        var votes = new Array(this.intercepts.length);
        for (var i = 0, d = 0, l = this.nRows; i < l; i++) {
            for (var j = i + 1; j < l; j++) {
                votes[d] = decisions[d] > 0 ? i : j;
                d++;
            }
        }
    
        var amounts = new Array(this.nClasses).fill(0);
        for (var i = 0, l = votes.length; i < l; i++) {
            amounts[votes[i]] += 1;
        }
    
        var classVal = -1, classIdx = -1;
        for (var i = 0, l = amounts.length; i < l; i++) {
            if (amounts[i] > classVal) {
                classVal = amounts[i];
                classIdx= i;
            }
        }
        return this.classes[classIdx];
    
    }

};

if (typeof process !== 'undefined' && typeof process.argv !== 'undefined') {
    if (process.argv.length - 2 === 6) {

        // Features:
        var features = process.argv.slice(2);

        // Parameters:
        var vectors = [[36.0, 31.0, 160.0, 166.0, 122.0, 137.0], [37.0, 31.0, 159.0, 165.0, 122.0, 137.0], [38.0, 27.0, 165.0, 171.0, 122.0, 136.0], [110.0, 23.0, 182.0, 178.0, 128.0, 121.0], [35.0, 29.0, 219.0, 221.0, 121.0, 140.0], [33.0, 51.0, 221.0, 222.0, 119.0, 149.0], [43.0, 21.0, 172.0, 177.0, 123.0, 134.0], [0.0, 0.0, 180.0, 187.0, 128.0, 128.0], [42.0, 22.0, 170.0, 175.0, 123.0, 134.0], [105.0, 11.0, 164.0, 168.0, 127.0, 125.0], [41.0, 42.0, 188.0, 191.0, 119.0, 142.0], [40.0, 25.0, 197.0, 200.0, 122.0, 137.0], [50.0, 33.0, 183.0, 185.0, 117.0, 138.0], [51.0, 34.0, 181.0, 183.0, 117.0, 138.0], [45.0, 35.0, 174.0, 177.0, 118.0, 139.0], [40.0, 40.0, 161.0, 166.0, 120.0, 140.0], [51.0, 32.0, 183.0, 185.0, 117.0, 138.0], [54.0, 27.0, 178.0, 181.0, 118.0, 136.0], [74.0, 24.0, 134.0, 140.0, 121.0, 130.0], [0.0, 0.0, 194.0, 200.0, 128.0, 128.0], [72.0, 34.0, 204.0, 203.0, 115.0, 133.0], [41.0, 29.0, 189.0, 192.0, 121.0, 138.0], [48.0, 34.0, 183.0, 185.0, 117.0, 138.0], [50.0, 33.0, 186.0, 188.0, 117.0, 138.0], [74.0, 24.0, 136.0, 142.0, 122.0, 130.0], [50.0, 34.0, 183.0, 185.0, 117.0, 138.0], [75.0, 27.0, 133.0, 138.0, 121.0, 130.0], [49.0, 37.0, 190.0, 191.0, 116.0, 139.0], [48.0, 36.0, 193.0, 194.0, 116.0, 140.0], [57.0, 27.0, 176.0, 179.0, 118.0, 135.0], [49.0, 34.0, 184.0, 186.0, 117.0, 138.0], [40.0, 44.0, 172.0, 176.0, 119.0, 142.0], [50.0, 31.0, 165.0, 169.0, 119.0, 137.0], [52.0, 31.0, 171.0, 174.0, 118.0, 137.0], [65.0, 33.0, 184.0, 185.0, 116.0, 134.0], [60.0, 25.0, 195.0, 196.0, 118.0, 135.0], [50.0, 34.0, 174.0, 177.0, 118.0, 138.0], [53.0, 32.0, 165.0, 168.0, 118.0, 137.0], [79.0, 44.0, 203.0, 202.0, 113.0, 130.0], [52.0, 33.0, 178.0, 180.0, 117.0, 138.0], [59.0, 30.0, 193.0, 194.0, 116.0, 137.0], [50.0, 36.0, 175.0, 177.0, 117.0, 138.0], [50.0, 33.0, 176.0, 179.0, 118.0, 137.0], [60.0, 27.0, 195.0, 196.0, 116.0, 136.0], [49.0, 34.0, 173.0, 176.0, 118.0, 138.0], [52.0, 35.0, 175.0, 177.0, 117.0, 138.0], [51.0, 33.0, 174.0, 177.0, 117.0, 137.0], [54.0, 32.0, 164.0, 167.0, 118.0, 136.0]];
        var coefficients = [[1.0, 0.5103328849993943, 1.0, 0.18460517859605727, 1.0, 0.0, 0.6794156464884626, 0.22175199029837298, 1.0, 0.0, -0.41406044750531895, -1.0, -0.0, -0.0, -0.0, -1.0, -0.0, -0.0, -0.0, -1.0, -0.0, -1.0, -0.0, -0.0, -1.0, -0.0, -0.0, -0.0, -0.0, -0.0, -0.0, -0.18204525287696996, -0.10220756665810597, -0.0, -0.0, -0.024832739041319236, -0.0, -0.0, -0.0, -0.0, -0.0, -0.0, -0.0, -0.0, -0.0, -0.0, -0.0, -0.0], [0.0, 0.0775378123102272, 0.0, 0.0, 0.0, 0.02096846829461621, 0.0, 0.0, 0.0, 0.028534025094581875, 0.0, 0.0, 1.0, 1.0, 1.0, 0.0, 1.0, 0.7629692919026765, 0.8615675910756857, 0.0, 1.0, 0.0, 1.0, 1.0, 0.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 0.0, -1.0, -1.0, -0.90466283661691, -1.0, -1.0, -1.0, -0.4956366212500156, -1.0, -1.0, -1.0, -1.0, -1.0, -1.0, -0.5075148620044435, -0.7167225631069957, -1.0]];
        var intercepts = [-384.0006853584271, -57.545497262241824, -277.21709072331515];
        var weights = [10, 22, 16];

        // Prediction:
        var clf = new SVC(3, 3, vectors, coefficients, intercepts, weights, "linear", 0.02, 0.0, 3);
        var prediction = clf.predict(features);
        console.log(prediction);

    }
}