Timeline for How can I identify correlated subqueries that can not be rewritten as JOINs without using DISTINCT?
Current License: CC BY-SA 3.0
11 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Jan 4, 2017 at 22:36 | comment | added | Evan Carroll | @user3686088 tried to make it clearer. Not 100% sure where I was going with that before. I think I was thinking about aggregates. | |
| Jan 4, 2017 at 22:36 | history | edited | Evan Carroll | CC BY-SA 3.0 |
deleted 86 characters in body
|
| Jan 4, 2017 at 4:46 | comment | added | RavBell | @EvanCaroll, Could you please clarify the first rule, particularly, I could not understand what you meant by "encapsulates only rows from one table." | |
| Jan 4, 2017 at 4:38 | vote | accept | RavBell | ||
| Dec 28, 2016 at 10:22 | comment | added | ypercubeᵀᴹ | I suggest you remove the "Moreover, rewriting the NOT IN as an anti-join (LEFT OUTER JOIN ... ON NULL) is likely faster." part. Unless you want to add an essay about efficiency, plans and query rewritings, to explain when this is true and when not ;) | |
| Dec 28, 2016 at 8:07 | comment | added | Lennart - Slava Ukraini |
Since the OP mentions Relational Algebra (RA) as a possible method, it might be worth noting that this will be difficult because RA is based on SET:s (where duplicates cannot exist by definition), whereas SQL is based on BAG:s.
|
|
| Dec 28, 2016 at 7:32 | history | edited | Evan Carroll | CC BY-SA 3.0 |
added 127 characters in body
|
| Dec 28, 2016 at 7:26 | history | edited | Evan Carroll | CC BY-SA 3.0 |
added 127 characters in body
|
| Dec 28, 2016 at 7:20 | history | edited | Evan Carroll | CC BY-SA 3.0 |
added 127 characters in body
|
| Dec 28, 2016 at 7:14 | history | edited | Evan Carroll | CC BY-SA 3.0 |
added 915 characters in body
|
| Dec 28, 2016 at 6:59 | history | answered | Evan Carroll | CC BY-SA 3.0 |