feat: set boxplot whisker as actual bound value(Q1 - N*IQR) or fixed value

common/changes/@visactor/vdataset/fix-box_plot_2025-11-18-07-47.json

@@ -0,0 +1,10 @@
+{
+  "changes": [
+    {
+      "packageName": "@visactor/vdataset",
+      "comment": "set boxplot whisker as actual bound value(Q1 - N*IQR) or fixed value && add alpha option of regression function",
+      "type": "none"
+    }
+  ],
+  "packageName": "@visactor/vdataset"
+}

packages/vdataset/__tests__/boxplot.test.ts

@@ -80,4 +80,74 @@ describe('boxplot transform', () => {
     expect(ids).toContain('a');
     expect(ids).toContain('b');
   });
+
+  test('whisker should be fixed not ranged by latset value', () => {
+    const data = [
+      {
+        v: -1
+      },
+      {
+        v: 3
+      },
+      {
+        v: 3
+      },
+      {
+        v: 3
+      },
+      {
+        v: 3
+      },
+      {
+        v: 3
+      },
+      {
+        v: 3
+      },
+      {
+        v: 3
+      },
+      {
+        v: 3
+      },
+      {
+        v: 3
+      },
+      {
+        v: 3
+      },
+      {
+        v: 3
+      },
+      {
+        v: 3
+      },
+      {
+        v: 3
+      },
+      {
+        v: 4
+      },
+      {
+        v: 4
+      },
+      {
+        v: 7
+      },
+      {
+        v: 8
+      },
+      {
+        v: 9
+      },
+      {
+        v: 10
+      }
+    ];
+    const out: any = boxplot(data, { field: 'v' });
+    expect(out.length).toBe(1);
+    const o = out[0];
+    expect(o.lowerWhisker).toBe(1.5);
+    expect(o.upperWhisker).toBe(5.5);
+  });
 });

packages/vdataset/src/transform/boxplot.ts

@@ -146,25 +146,9 @@ export const boxplot: Transform = (data: Array<object>, options?: IBoxplotOption
     const lowerBound = isArray(whiskers) ? quantileSorted(sorted, whiskers[0]) : q1 - whiskers * iqr;
     const upperBound = isArray(whiskers) ? quantileSorted(sorted, whiskers[1]) : q3 + whiskers * iqr;
 
-    // whiskers are the most extreme data points inside the bounds
-    let lowerWhisker = dataMin;
-    for (let i = 0; i < sorted.length; i++) {
-      if (sorted[i] >= lowerBound) {
-        lowerWhisker = sorted[i];
-        break;
-      }
-    }
-    let upperWhisker = dataMax;
-    for (let i = sorted.length - 1; i >= 0; i--) {
-      if (sorted[i] <= upperBound) {
-        upperWhisker = sorted[i];
-        break;
-      }
-    }
-
     const outliers: number[] = [];
     for (let i = 0; i < sorted.length; i++) {
-      if (sorted[i] < lowerWhisker || sorted[i] > upperWhisker) {
+      if (sorted[i] < lowerBound || sorted[i] > upperBound) {
         outliers.push(sorted[i]);
       }
     }
@@ -191,8 +175,8 @@ export const boxplot: Transform = (data: Array<object>, options?: IBoxplotOption
     obj[iqrName] = iqr;
     obj[minName] = dataMin;
     obj[maxName] = dataMax;
-    obj[lowerWhiskerName] = lowerWhisker;
-    obj[upperWhiskerName] = upperWhisker;
+    obj[lowerWhiskerName] = lowerBound;
+    obj[upperWhiskerName] = upperBound;
     obj[outliersName] = outliers;
     if (includeValues) {
       obj[valuesName] = rawValues.get(key) || [];

packages/vutils/src/common/regression-linear.ts

@@ -55,7 +55,15 @@ export function rSquared(
   return sst === 0 ? 0 : 1 - ssr / sst;
 }
 
-export function regressionLinear(data: any[], x: (d: any) => number = d => d.x, y: (d: any) => number = d => d.y) {
+export function regressionLinear(
+  data: any[],
+  x: (d: any) => number = d => d.x,
+  y: (d: any) => number = d => d.y,
+  options?: {
+    alpha?: number;
+  }
+) {
+  const alpha = options?.alpha ?? 0.05;
   // accumulate online means (sufficient statistics)
   let n = 0;
   let meanX = 0;
@@ -95,7 +103,7 @@ export function regressionLinear(data: any[], x: (d: any) => number = d => d.x,
     return out;
   }
 
-  function confidenceInterval(N: number = 50, alpha: number = 0.05) {
+  function confidenceInterval(N: number = 50) {
     const out: { x: number; mean: number; lower: number; upper: number; predLower: number; predUpper: number }[] = [];
     if (comps.n === 0 || N <= 0) {
       return out;

packages/vutils/src/common/regression-logistic.ts

@@ -11,10 +11,11 @@ export function regressionLogistic(
   data: any[],
   x: (d: any) => number = d => d.x,
   y: (d: any) => number = d => d.y,
-  options?: { maxIteration?: number; tol?: number }
+  options?: { maxIteration?: number; tol?: number; alpha?: number }
 ) {
   const maxIter = options?.maxIteration ?? 25;
   const tol = options?.tol ?? 1e-6;
+  const alpha = options?.alpha ?? 0.05;
   // build arrays
   const xs: number[] = [];
   const ys: number[] = [];
@@ -120,7 +121,7 @@ export function regressionLogistic(
     return out;
   }
 
-  function confidenceInterval(N: number = 50, alpha: number = 0.05) {
+  function confidenceInterval(N: number = 50) {
     const out: { x: number; mean: number; lower: number; upper: number; predLower: number; predUpper: number }[] = [];
 
     if (N <= 0) {

packages/vutils/src/common/regression-lowess.ts

@@ -21,10 +21,11 @@ export function regressionLowess(
   data: any[],
   x: (d: any) => number = d => d.x,
   y: (d: any) => number = d => d.y,
-  options: { span?: number; degree?: 1 | 0; iterations?: number } = {}
+  options: { span?: number; degree?: 1 | 0; iterations?: number; alpha?: number } = {}
 ) {
   const span = options.span || 0.3;
   const degree = options.degree === 0 ? 0 : 1;
+  const alpha = options?.alpha ?? 0.05;
   const iterations = options.iterations == null ? 2 : options.iterations;
 
   const ptsX: number[] = [];
@@ -170,7 +171,7 @@ export function regressionLowess(
     return out;
   }
 
-  function confidenceInterval(N: number = 50, alpha: number = 0.05) {
+  function confidenceInterval(N: number = 50) {
     const out: { x: number; mean: number; lower: number; upper: number; predLower: number; predUpper: number }[] = [];
 
     if (N <= 0) {

packages/vutils/src/common/regression-polynomial.ts

@@ -69,12 +69,13 @@ export function regressionPolynomial(
   data: any[],
   x: (d: any) => number = d => d.x,
   y: (d: any) => number = d => d.y,
-  options: { degree?: number } = {}
+  options: { degree?: number; alpah?: number } = {}
 ) {
   let degree = options.degree ?? 0;
   if (degree < 0) {
     degree = 0;
   }
+  const alpha = options.alpah ?? 0.5;
   const m = degree + 1;
   const sums: number[] = new Array(2 * degree + 1).fill(0);
 
@@ -171,7 +172,7 @@ export function regressionPolynomial(
       }
       return out;
     },
-    confidenceInterval(N: number = 50, alpha: number = 0.05) {
+    confidenceInterval(N: number = 50) {
       const out: { x: number; mean: number; lower: number; upper: number; predLower: number; predUpper: number }[] = [];
 
       if (N <= 0) {