[SPARK-37165][SQL] Add REPEATABLE in TABLESAMPLE to specify seed #34442
Conversation
|
Kubernetes integration test starting |
|
Kubernetes integration test status failure |
|
cc @viirya |
|
Test build #144761 has finished for PR 34442 at commit
|
|
retest this please |
|
Kubernetes integration test starting |
|
Kubernetes integration test status failure |
| * are defined as a number between 0 and 100. | ||
| * - TABLESAMPLE(BUCKET x OUT OF y): Sample the table down to a 'x' divided by 'y' fraction. | ||
| */ | ||
| private def withSample(ctx: SampleContext, query: LogicalPlan): LogicalPlan = withOrigin(ctx) { |
There was a problem hiding this comment.
We should also update method doc.
| test("TABLE SAMPLE") { | ||
| withTable("test") { | ||
| sql("CREATE TABLE test(c int) USING PARQUET") | ||
| for( i <- 0 to 20) { |
viirya
left a comment
viirya
left a comment
There was a problem hiding this comment.
Looks okay. Just a few minor comments.
|
Test build #144769 has finished for PR 34442 at commit
|
|
Kubernetes integration test starting |
|
Kubernetes integration test status failure |
|
Test build #144775 has finished for PR 34442 at commit
|
|
Merged to mater. Thanks a lot for reviewing! @viirya |
|
FYI @cloud-fan |
|
late LGTM |
|
Was this merged in the end? I can only see the PR as closed (not merged) and I can't see the functionality in the spark SQL docs or the codebase - but aware I may be missing something! |
|
@RossKen This PR was merged. |
5 participants
What changes were proposed in this pull request?
Add REPEATABLE in SQL syntax TABLESAMPLE so user can specify seed.
Why are the changes needed?
Current syntax for TABLESAMPLE:
Dataset.samplehas a param to specify seed, so we should allow SQL has a way to specify seed too.Most of the DBMS uses REPEATABLE to let user specify seed, e.g. DB2, we will follow the same way.
Does this PR introduce any user-facing change?
Yes
new SQL syntax
How was this patch tested?
new UT