CODEWITHSUNDEEP
teradatadata-partitioning
n8.
n8.

Reputation: 1738

Top or Sample N of subgroups in Teradata in a large data set ("No Spool Space" error)

I've tried several routes to getting the 10 records from each subset of a large dataset and the best I can do is querying each subgroup explicitly in the query.

My first attempt from the (Teradata community help website) was this:

Select * FROM Table Qualify Row_number() Over (partition by date order by column) between 1 and 10;

I wanted this to work because I didn't want to manually go and harvest the partitions, but I run out of spool space when I try that. In the same forum there's this solution


  FROM TheTable

SAMPLE WHEN TheDate = '2014-01-01' THEN 10

            WHEN TheDate = '2014-2-01' THEN 10

...

END;

It works, but obviously is not dynamic to the dataset contents. Can I get a solution that is dynamic for large data sets?

Upvotes: 0

Views: 23

Answers (0)

Related Questions