@ai-on-browser/data-analysis-models/lib/model/cubic_hermite_spline.js

Data analysis model package without any dependencies

/**
 * Cubic Hermite spline
 */
export default class CubicHermiteSpline {
	// http://paulbourke.net/miscellaneous/interpolation/
	// https://en.wikipedia.org/wiki/Cubic_Hermite_spline
	/**
	 * @param {number} t Tension factor
	 * @param {number} b Bias factor
	 */
	constructor(t, b) {
		this._t = t
		this._b = b
	}

	/**
	 * Fit model parameters.
	 * @param {number[]} x Training data
	 * @param {number[]} y Target values
	 */
	fit(x, y) {
		const d = x.map((v, i) => [v, y[i]])
		d.sort((a, b) => a[0] - b[0])
		this._x = d.map(v => v[0])
		this._y = d.map(v => v[1])
	}

	/**
	 * Returns predicted interpolated values.
	 * @param {number[]} target Sample data
	 * @returns {number[]} Predicted values
	 */
	predict(target) {
		const n = this._x.length
		return target.map(t => {
			if (t <= this._x[0]) {
				return this._y[0]
			} else if (t >= this._x[n - 1]) {
				return this._y[n - 1]
			}
			for (let i = 1; i < n; i++) {
				if (t <= this._x[i]) {
					const p = (t - this._x[i - 1]) / (this._x[i] - this._x[i - 1])
					const y0 = i > 1 ? this._y[i - 2] : 2 * this._y[i - 1] - this._y[i]
					const y1 = this._y[i - 1]
					const y2 = this._y[i]
					const y3 = i < n - 1 ? this._y[i + 1] : 2 * this._y[i] + this._y[i - 1]
					const m0 = (((y1 - y0) * (1 + this._b) + (y2 - y1) * (1 - this._b)) * (1 - this._t)) / 2
					const m1 = (((y2 - y1) * (1 + this._b) + (y3 - y2) * (1 - this._b)) * (1 - this._t)) / 2
					const a0 = 2 * p ** 3 - 3 * p ** 2 + 1
					const a1 = p ** 3 - 2 * p ** 2 + p
					const a2 = p ** 3 - p ** 2
					const a3 = -2 * p ** 3 + 3 * p ** 2
					return a0 * y1 + a1 * m0 + a2 * m1 + a3 * y2
				}
			}
			return this._y[n - 1]
		})
	}
}