Using numeric data types

Hive supports a set of data types that can be used for table columns, expression values, and function arguments, and return values.

In the following table, primitive numeric data types are listed with sizes and examples:

Data Type	Size	Example
TINYINT	1-byte signed integer	50
SMALLINT	2-byte signed integer	20,000
INT	4-byte signed integer	1,000
BIGINT	8-byte signed integer	50,000
FLOAT	4-byte single-precision floating point	400.50
DOUBLE	8-byte double-precision floating point	20,000.50
DECIMAL	17-byte precision up to 38 digits	DECIMAL(20,2)

By default, all integral literals are treated as the INT values until they cross the range of INT values. If some integral literal crosses the range of the INT value, then it is treated as the BIGINT value. There is a mechanism of postfix, which is used to specify an integral literal as TINYINT, SMALLINT, and BIGINT.

To specify an integral literal as TINYINT, the postfix Y is used. For example, you can specify...