Numeric Precision Parameters

The following configuration parameters let you configure numeric precision for numeric data types. For more about using these parameters, see Numeric Data Type Overflow with SUM, SUM_FLOAT, and AVG.

Parameters Description

AllowNumericOverflow

When set to 1 (True), allows silent numeric overflow. When true, Vertica does not implicitly extend precision of numeric data types.

Default Value: 1 (True)

When set to 0 (False), Vertica produces an overflow error, if a result exceeds the precision set by NumericSumExtraPrecisionDigits.

Vertica ignores the value of NumericSumExtraPrecisionDigits when AllowNumericOverflow is true.

Example:

ALTER DATABASE mydb SET PARAMETER AllowNumericOverflow=0;

NumericSumExtraPrecisionDigits

Vertica produces an overflow error, if a result exceeds the specified precision, By default this specified precision is six places beyond the DDL-specified precision.

Default Value: 6

This parameter setting only applies if AllowNumericOverflow is set to 0 (False).

Possible Values: 0 to 20.

Example:

ALTER DATABASE mydb SET PARAMETER NumericSumExtraPrecisionDigits=8