ALTER DATASET … ADD VERSION

Adds a version to a dataset. When you add a version, you can specify properties such as partitioning, comments, or custom metadata.

See also:

ALTER DATASET , ALTER DATASET … DROP VERSION

Syntax

ALTER DATASET <name> ADD VERSION <version_name>
  FROM <select_statement>
  [ PARTITION BY <string_expr> ]
  [ COMMENT = <string_literal> ]
  [ METADATA = <json_string_literal> ]
Copy

Parameters

name

The name of the dataset that you’re altering.

ADD VERSION version_name

The name of the new dataset version that you’re creating.

FROM select_statement

The SQL statement that defines the data for the new dataset version.

PARTITION BY string_expr

The partitioning expression for the new dataset version.

COMMENT = string_literal

A comment for the new dataset version.

METADATA = json_string_literal

A JSON string containing metadata for the new dataset version. The following is an example of a JSON string.

{"source": "my_table", "job_id": "123"}
Copy

Access control requirements

Privilege

Object

Notes

OWNERSHIP

Schema

Provides the privilege to both read and modify the dataset.

Examples

The following example adds version v1 to the abc dataset with partitioning:

ALTER DATASET abc
ADD VERSION 'v1' FROM (
    SELECT seq4() as ID, uniform(1, 10, random(721)) as PART
    FROM TABLE(GENERATOR(ROWCOUNT => 100000)) v)
PARTITION BY PART
COMMENT = 'Initial version'
METADATA = '{"source":"some_table","created_by":"analyst1"}';
Copy
Language: English