#weightedMean

Calculates a weighted arithmetic mean without requiring weights to sum to 1.

import { Arith } from "@teakit/arith";

Arith.weightedMean(["10", "20", "30"], ["0.2", "0.3", "0.5"]).toString(); // "23"
Arith.weightedMean(["100", "120", "80"], ["5", "3", "2"]).toString(); // "102"

// Zero weights are allowed if at least one weight is positive.
Arith.weightedMean(["10", "20", "30"], [0, 1, 0]).toString(); // "20"

const Short = Arith.clone({ DECIMAL_PLACES: 4 });
Short.weightedMean(["1", "2"], ["1", "2"]).toString(); // "1.6667"

#API Reference

#Signature

Arith.weightedMean(
  values: readonly ArithValue[],
  weights: readonly ArithValue[],
): ArithInstance;

#Parameters

#Returns

Returns sum(value[i] * weight[i]) / sum(weights) as a new Arith instance.

#Throws

• Throws if values or weights is not an array.
• Throws if their lengths differ.
• Throws if any value is non-finite.
• Throws if any weight is non-finite or negative.
• Throws if all weights are zero.

#Agent Contract

#Agent Notes

• Weights do not need to be normalized.
• Use weightedMean for weighted cost, portfolio weight, or score aggregation.
• Do not generate average or other aliases.