Snowpark Migration Accelerator: Spark SQL Data Types¶

Conversion Table¶

Spark SQL	Snowflake	Notes
`BIGINT`	`BIGINT`
`BOOLEAN`	`BOOLEAN`
`BYTE`	`BYTEINT`
`CHAR`	`CHAR`
`DATE`	`DATE`
`DECIMAL`	`DECIMAL`
`DOUBLE`	`DOUBLE`
`FLOAT`	`FLOAT`
`INTEGER`	`INTEGER`
`LONG`	`INT`	Check out note
`SHORT`	`INT`	Check out note
`STRING`	`STRING`
`TIMESTAMP`	`TIMESTAMP_TZ`
`TIMESTAMPNTZ`	`TIMESTAMP_NTZ`
`VARCHAR`	`VARCHAR`

Notes¶

Note

For more information, refer to the Spark SQL data types documentation.

Integer types¶

When converting integer data types from the source system, both LONG and SHORT are mapped to Snowflake’s INTEGER data type, as INTEGER can accommodate the full range of values for both data types.

SparkSQL LONG: Range from -32,768 to 32,767
SparkSQL SHORT: Range from -9,223,372,036,854,775,808 to 9,223,372,036,854,775,807
Snowflake INTEGER: Range from -9.9999999999999999999999999999999999999 x 10^38 to +9.9999999999999999999999999999999999999 x 10^38