Open billhowe opened 9 years ago
To be clear, are you proposing to:
1) add syntax to MyriaL for a left-join operator
2) add such an operator to LogicalAlgebra
On Mon, Apr 6, 2015 at 10:40 AM, Brandon Myers notifications@github.com wrote:
To be clear, you are proposing to add syntax to MyriaL for a left-join operator and add such an operator to LogicalAlgebra?
I think so, yes.
Bill Howe Associate Director and Senior Data Science Fellow, UW eScience Institute Affiliate Faculty, Computer Science & Engineering University of Washington To acknowledge eScience: "Supported in part by the University of Washington eScience Institute"
Left or right? Myria has a right hash join ;-)
I also have a use-case where it would be useful to surface a left/right join at the MyriaL level.
Left-only is ok with me. I find people expect the "root" relation to be on the left typically.
On Tue, Apr 7, 2015 at 7:14 PM, Dominik Moritz notifications@github.com wrote:
Left or right? Myria has a right hash join https://github.com/uwescience/myria/blob/master/src/edu/washington/escience/myria/operator/RightHashJoin.java ;-)
— Reply to this email directly or view it on GitHub https://github.com/uwescience/raco/issues/402#issuecomment-90781793.
Bill Howe Associate Director and Senior Data Science Fellow, UW eScience Institute Affiliate Faculty, Computer Science & Engineering University of Washington To acknowledge eScience: "Supported in part by the University of Washington eScience Institute"
Unless we have it in some form already.
Sp2Bench is the immediate use case.
We can fake it with inner join unioned with the stuff that fails to join, but that's a terrifyingly large amount of work compared to just emitting the left tuple padded with nulls instead of dropping it on the floor.