-
Notifications
You must be signed in to change notification settings - Fork 9
Closed
Description
Background
Floating-point numbers no always have exact representation, so a rounding error can be introduced when summing up floating-point numbers.
In order to control accuracy and completeness of such numbers it makes sense to truncate the fractional part, summing up only the integral part of numbers, which have exact representation.
Feature
Add support for sum of truncated values.
Proposed solusion
| Type | Description |
|---|---|
| controlType.aggregatedTruncTotal | Calculates SUM(TRUNC()) of the specified column |
| controlType.absAggregatedTruncTotal | Calculates SUM(TRUNC(ABS())) of the specified column |
SUM(TRUNC(x)) - Sum of truncated numbers of a column x.
SUM(TRUNC(ABS(x)) - Sum of truncated absolute values numbers of a column x.
Examples
trunc(123) = 123trunc(123.0001) = 123trunc(123.456) = 123trunc(123.999) = 123trunc(-123.456) = -123trunc(-123.999) = -123
Metadata
Metadata
Assignees
Labels
No labels