wouterbulten/slacjs

/**
 * Random following normal distribution
 * @param  {float} mean mean
 * @param  {float} sd   standard deviation
 * @return {float}
 */
export function randn(mean, sd) {

    //Retrieved from jStat
    let u;
    let v;
    let x;
    let y;
    let q;

    do {
        u = Math.random();
        v = 1.7156 * (Math.random() - 0.5);
        x = u - 0.449871;
        y = Math.abs(v) + 0.386595;
        q = x * x + y * (0.19600 * y - 0.25472 * x);
    } while (q > 0.27597 && (q > 0.27846 || v * v > -4 * Math.log(u) * u * u));

    return (v / u) * sd + mean;
}

/**
 * pdf for a normal distribution
 * @param  {Number} x
 * @param  {Number} mean
 * @param  {Number} sd
 * @return {Number}
 */
export function pdfn(x, mean, sd) {
    return (1 / (sd * Math.sqrt(2 * Math.PI))) * Math.exp(-(Math.pow(x - mean, 2)) / (2 * sd * sd));
}

/**
 * Compute the log with a given base
 *
 * Used primarily as log10 is not implemented yet on mobile browsers
 * @param  {int}
 * @param  {int}
 * @return {float}
 */
export function log(x, base) {
    return Math.log(x) / Math.log(base);
}

/**
 * Calculates two eigenvalues and eigenvectors from a 2x2 covariance matrix
 * @param  {Array} cov
 * @return {object}
 */
export function eigenvv(cov) {

    const a = cov[0][0];
    const b = cov[0][1];
    const c = cov[1][0];
    const d = cov[1][1];

    const A = 1;
    const B = -(a + d);

    //const C = (a * d) - (c * b);

    const L1 = (-B + Math.sqrt((Math.pow(a - d, 2) + (4 * c * d))) / 2 * A);
    const L2 = (-B - Math.sqrt((Math.pow(a - d, 2) + (4 * c * d))) / 2 * A);

    const y1 = (L1 - a) / b;
    const y2 = (L2 - a) / b;
    const mag1 = Math.sqrt(1 + (y1 * y1));
    const mag2 = Math.sqrt(1 + (y2 * y2));

    return {
        values: [L1, L2],
        vectors: [[1 / mag1, y1 / mag1], [1 / mag2, y2 / mag2]]
    };
}

/**
 * Calculate the variance of an array given a value function
 * @param  {Array} data
 * @param  {Function} valueFunc Function that maps an array element to a number
 * @return {Number}
 */
export function variance(data, valueFunc) {

    let sum = 0;
    let sumSq = 0;
    const n = data.length;

    data.forEach((d) => {

        const value = valueFunc(d);

        sum += value;
        sumSq += (value * value);
    });

    return (sumSq - ((sum * sum) / n)) / n;
}