ddSum
ddSum(qs: number[][]): number[]
Defined in double-double/multi/dd-sum.ts:12
Returns the result of summing an array of double-double-precision floating point numbers naively (i.e. not using pairwise addition to reduce error a bit).
- an error bound is given by: (n-1)(1+ϵ), where ϵ <= 3u^2 + 13u^3, u = 0.5 * Number.EPSILON
Parameters:
| Name | Type |
|---|---|
qs | number[][] |
Returns: number[]