feat: add mean absolute deviation algorithm

Maths/MeanAbsoluteDeviation.js

@@ -1,21 +1,23 @@
-import { mean } from './AverageMean.js'
 /**
- *@function meanAbsoluteDeviation
- *@description Calculates the mean absolute deviation of list of numbers
- * @param {Integer} data
- * @returns meanAbsoluteDeviation([2,34,5,0,-2]) = 10.480
- * @url https://en.wikipedia.org/wiki/Average_absolute_deviation
+ * @function meanAbsoluteDeviation
+ * @description Calculates the mean absolute deviation (MAD) of an array
+ *   of numbers. MAD is the average of the absolute differences between
+ *   each value and the mean of the dataset.
+ * @param {number[]} arr - An array of numbers
+ * @returns {number} - The mean absolute deviation
+ * @see https://en.wikipedia.org/wiki/Average_absolute_deviation
+ * @example meanAbsoluteDeviation([2, 2, 3, 4, 14]) => 3.6
+ * @example meanAbsoluteDeviation([1, 1, 1]) => 0
  */
-function meanAbsoluteDeviation(data) {
-  if (!Array.isArray(data)) {
-    throw new TypeError('Invalid Input')
+const meanAbsoluteDeviation = (arr) => {
+  if (!Array.isArray(arr) || arr.length === 0) {
+    throw new TypeError('Input must be a non-empty array')
   }
-  let absoluteSum = 0
-  const meanValue = mean(data)
-  for (const dataPoint of data) {
-    absoluteSum += Math.abs(dataPoint - meanValue)
+  if (!arr.every((n) => typeof n === 'number' && !isNaN(n))) {
+    throw new TypeError('All elements must be numbers')
   }
-  return absoluteSum / data.length
+  const mean = arr.reduce((sum, n) => sum + n, 0) / arr.length
+  return arr.reduce((sum, n) => sum + Math.abs(n - mean), 0) / arr.length
 }
 
 export { meanAbsoluteDeviation }

Maths/test/MeanAbsoluteDeviation.test.js

@@ -1,16 +1,45 @@
 import { meanAbsoluteDeviation } from '../MeanAbsoluteDeviation.js'
 
-describe('tests for mean absolute deviation', () => {
-  it('should be a function', () => {
-    expect(typeof meanAbsoluteDeviation).toEqual('function')
+describe('meanAbsoluteDeviation', () => {
+  test('meanAbsoluteDeviation([2,2,3,4,14]) returns 3.6', () => {
+    expect(meanAbsoluteDeviation([2, 2, 3, 4, 14])).toBeCloseTo(3.6)
   })
 
-  it('should throw an invalid input error', () => {
-    expect(() => meanAbsoluteDeviation('fgh')).toThrow()
+  test('meanAbsoluteDeviation([1,1,1]) returns 0', () => {
+    expect(meanAbsoluteDeviation([1, 1, 1])).toBe(0)
   })
 
-  it('should return the mean absolute deviation of an array of numbers', () => {
-    const meanAbDev = meanAbsoluteDeviation([2, 34, 5, 0, -2])
-    expect(meanAbDev).toBe(10.479999999999999)
+  test('meanAbsoluteDeviation single element returns 0', () => {
+    expect(meanAbsoluteDeviation([5])).toBe(0)
+  })
+
+  test('meanAbsoluteDeviation([1,2,3,4,5]) returns 1.2', () => {
+    expect(meanAbsoluteDeviation([1, 2, 3, 4, 5])).toBeCloseTo(1.2)
+  })
+
+  test('meanAbsoluteDeviation works with negative numbers', () => {
+    expect(meanAbsoluteDeviation([-2, -1, 0, 1, 2])).toBeCloseTo(1.2)
+  })
+
+  test('meanAbsoluteDeviation([10, 20]) returns 5', () => {
+    expect(meanAbsoluteDeviation([10, 20])).toBeCloseTo(5)
+  })
+
+  test('meanAbsoluteDeviation does not mutate original array', () => {
+    const arr = [1, 2, 3, 4, 5]
+    meanAbsoluteDeviation(arr)
+    expect(arr).toEqual([1, 2, 3, 4, 5])
+  })
+
+  test('throws TypeError for empty array', () => {
+    expect(() => meanAbsoluteDeviation([])).toThrow(TypeError)
+  })
+
+  test('throws TypeError for non-array input', () => {
+    expect(() => meanAbsoluteDeviation('123')).toThrow(TypeError)
+  })
+
+  test('throws TypeError for array with non-numbers', () => {
+    expect(() => meanAbsoluteDeviation([1, 'two', 3])).toThrow(TypeError)
   })
 })