chartjs-plugin-trendline 3.2.0 → 3.2.3

package/src/utils/exponentialFitter.js

@@ -1,147 +1,147 @@
-/**
- * A class that fits an exponential curve to a series of points using least squares.
- * Fits y = a * e^(b*x) by transforming to ln(y) = ln(a) + b*x
- */
-export class ExponentialFitter {
-    constructor() {
-        this.count = 0;
-        this.sumx = 0;
-        this.sumlny = 0;
-        this.sumx2 = 0;
-        this.sumxlny = 0;
-        this.minx = Number.MAX_VALUE;
-        this.maxx = Number.MIN_VALUE;
-        this.hasValidData = true;
-        this.dataPoints = []; // Store data points for correlation calculation
-        this._cachedGrowthRate = null;
-        this._cachedCoefficient = null;
-        this._cachedCorrelation = null;
-        this._cacheValid = false;
-    }
-
-    /**
-     * Adds a point to the exponential fitter.
-     * @param {number} x - The x-coordinate of the point.
-     * @param {number} y - The y-coordinate of the point.
-     */
-    add(x, y) {
-        if (y <= 0) {
-            this.hasValidData = false;
-            return;
-        }
-
-        const lny = Math.log(y);
-        if (!isFinite(lny)) {
-            this.hasValidData = false;
-            return;
-        }
-
-        this.sumx += x;
-        this.sumlny += lny;
-        this.sumx2 += x * x;
-        this.sumxlny += x * lny;
-        if (x < this.minx) this.minx = x;
-        if (x > this.maxx) this.maxx = x;
-        this.dataPoints.push({x, y, lny}); // Store actual data points
-        this.count++;
-        this._cacheValid = false;
-    }
-
-    /**
-     * Calculates the exponential growth rate (b in y = a * e^(b*x)).
-     * @returns {number} - The exponential growth rate.
-     */
-    growthRate() {
-        if (!this.hasValidData || this.count < 2) return 0;
-        if (!this._cacheValid) {
-            this._computeCoefficients();
-        }
-        return this._cachedGrowthRate;
-    }
-
-    /**
-     * Calculates the exponential coefficient (a in y = a * e^(b*x)).
-     * @returns {number} - The exponential coefficient.
-     */
-    coefficient() {
-        if (!this.hasValidData || this.count < 2) return 1;
-        if (!this._cacheValid) {
-            this._computeCoefficients();
-        }
-        return this._cachedCoefficient;
-    }
-
-    /**
-     * Returns the fitted exponential value (y) for a given x.
-     * @param {number} x - The x-coordinate.
-     * @returns {number} - The corresponding y-coordinate on the fitted exponential curve.
-     */
-    f(x) {
-        if (!this.hasValidData || this.count < 2) return 0;
-        if (!this._cacheValid) {
-            this._computeCoefficients();
-        }
-        
-        // Check for potential overflow before calculation
-        if (Math.abs(this._cachedGrowthRate * x) > 500) return 0; // Safer limit to prevent overflow
-        
-        const result = this._cachedCoefficient * Math.exp(this._cachedGrowthRate * x);
-        return isFinite(result) ? result : 0;
-    }
-
-    /**
-     * Calculates the correlation coefficient (R-squared) for the exponential fit.
-     * @returns {number} - The correlation coefficient (0-1).
-     */
-    correlation() {
-        if (!this.hasValidData || this.count < 2) return 0;
-        if (!this._cacheValid) {
-            this._computeCoefficients();
-        }
-        return this._cachedCorrelation;
-    }
-
-    /**
-     * Returns the scale (growth rate) of the fitted exponential curve.
-     * @returns {number} - The growth rate of the exponential curve.
-     */
-    scale() {
-        return this.growthRate();
-    }
-
-    _computeCoefficients() {
-        if (!this.hasValidData || this.count < 2) {
-            this._cachedGrowthRate = 0;
-            this._cachedCoefficient = 1;
-            this._cachedCorrelation = 0;
-            this._cacheValid = true;
-            return;
-        }
-
-        const denominator = this.count * this.sumx2 - this.sumx * this.sumx;
-        if (Math.abs(denominator) < 1e-10) {
-            this._cachedGrowthRate = 0;
-            this._cachedCoefficient = 1;
-            this._cachedCorrelation = 0;
-            this._cacheValid = true;
-            return;
-        }
-
-        this._cachedGrowthRate = (this.count * this.sumxlny - this.sumx * this.sumlny) / denominator;
-        const lnA = (this.sumlny - this._cachedGrowthRate * this.sumx) / this.count;
-        this._cachedCoefficient = Math.exp(lnA);
-
-        const meanLnY = this.sumlny / this.count;
-        let ssTotal = 0;
-        let ssRes = 0;
-        
-        for (const point of this.dataPoints) {
-            const predictedLnY = lnA + this._cachedGrowthRate * point.x;
-            ssTotal += Math.pow(point.lny - meanLnY, 2);
-            ssRes += Math.pow(point.lny - predictedLnY, 2);
-        }
-        
-        this._cachedCorrelation = ssTotal === 0 ? 1 : Math.max(0, 1 - (ssRes / ssTotal));
-        this._cacheValid = true;
-    }
+/**
+ * A class that fits an exponential curve to a series of points using least squares.
+ * Fits y = a * e^(b*x) by transforming to ln(y) = ln(a) + b*x
+ */
+export class ExponentialFitter {
+    constructor() {
+        this.count = 0;
+        this.sumx = 0;
+        this.sumlny = 0;
+        this.sumx2 = 0;
+        this.sumxlny = 0;
+        this.minx = Number.MAX_VALUE;
+        this.maxx = Number.MIN_VALUE;
+        this.hasValidData = true;
+        this.dataPoints = []; // Store data points for correlation calculation
+        this._cachedGrowthRate = null;
+        this._cachedCoefficient = null;
+        this._cachedCorrelation = null;
+        this._cacheValid = false;
+    }
+
+    /**
+     * Adds a point to the exponential fitter.
+     * @param {number} x - The x-coordinate of the point.
+     * @param {number} y - The y-coordinate of the point.
+     */
+    add(x, y) {
+        if (y <= 0) {
+            this.hasValidData = false;
+            return;
+        }
+
+        const lny = Math.log(y);
+        if (!isFinite(lny)) {
+            this.hasValidData = false;
+            return;
+        }
+
+        this.sumx += x;
+        this.sumlny += lny;
+        this.sumx2 += x * x;
+        this.sumxlny += x * lny;
+        if (x < this.minx) this.minx = x;
+        if (x > this.maxx) this.maxx = x;
+        this.dataPoints.push({x, y, lny}); // Store actual data points
+        this.count++;
+        this._cacheValid = false;
+    }
+
+    /**
+     * Calculates the exponential growth rate (b in y = a * e^(b*x)).
+     * @returns {number} - The exponential growth rate.
+     */
+    growthRate() {
+        if (!this.hasValidData || this.count < 2) return 0;
+        if (!this._cacheValid) {
+            this._computeCoefficients();
+        }
+        return this._cachedGrowthRate;
+    }
+
+    /**
+     * Calculates the exponential coefficient (a in y = a * e^(b*x)).
+     * @returns {number} - The exponential coefficient.
+     */
+    coefficient() {
+        if (!this.hasValidData || this.count < 2) return 1;
+        if (!this._cacheValid) {
+            this._computeCoefficients();
+        }
+        return this._cachedCoefficient;
+    }
+
+    /**
+     * Returns the fitted exponential value (y) for a given x.
+     * @param {number} x - The x-coordinate.
+     * @returns {number} - The corresponding y-coordinate on the fitted exponential curve.
+     */
+    f(x) {
+        if (!this.hasValidData || this.count < 2) return 0;
+        if (!this._cacheValid) {
+            this._computeCoefficients();
+        }
+        
+        // Check for potential overflow before calculation
+        if (Math.abs(this._cachedGrowthRate * x) > 500) return 0; // Safer limit to prevent overflow
+        
+        const result = this._cachedCoefficient * Math.exp(this._cachedGrowthRate * x);
+        return isFinite(result) ? result : 0;
+    }
+
+    /**
+     * Calculates the correlation coefficient (R-squared) for the exponential fit.
+     * @returns {number} - The correlation coefficient (0-1).
+     */
+    correlation() {
+        if (!this.hasValidData || this.count < 2) return 0;
+        if (!this._cacheValid) {
+            this._computeCoefficients();
+        }
+        return this._cachedCorrelation;
+    }
+
+    /**
+     * Returns the scale (growth rate) of the fitted exponential curve.
+     * @returns {number} - The growth rate of the exponential curve.
+     */
+    scale() {
+        return this.growthRate();
+    }
+
+    _computeCoefficients() {
+        if (!this.hasValidData || this.count < 2) {
+            this._cachedGrowthRate = 0;
+            this._cachedCoefficient = 1;
+            this._cachedCorrelation = 0;
+            this._cacheValid = true;
+            return;
+        }
+
+        const denominator = this.count * this.sumx2 - this.sumx * this.sumx;
+        if (Math.abs(denominator) < 1e-10) {
+            this._cachedGrowthRate = 0;
+            this._cachedCoefficient = 1;
+            this._cachedCorrelation = 0;
+            this._cacheValid = true;
+            return;
+        }
+
+        this._cachedGrowthRate = (this.count * this.sumxlny - this.sumx * this.sumlny) / denominator;
+        const lnA = (this.sumlny - this._cachedGrowthRate * this.sumx) / this.count;
+        this._cachedCoefficient = Math.exp(lnA);
+
+        const meanLnY = this.sumlny / this.count;
+        let ssTotal = 0;
+        let ssRes = 0;
+        
+        for (const point of this.dataPoints) {
+            const predictedLnY = lnA + this._cachedGrowthRate * point.x;
+            ssTotal += Math.pow(point.lny - meanLnY, 2);
+            ssRes += Math.pow(point.lny - predictedLnY, 2);
+        }
+        
+        this._cachedCorrelation = ssTotal === 0 ? 1 : Math.max(0, 1 - (ssRes / ssTotal));
+        this._cacheValid = true;
+    }
 }