Detect columns with dependencies and suggest to create extended statistics on it

[sfkeller]

Detect correlated columns (i.e. columns with dependencies) based on slow queries.
Then suggest a CREATE STATISTICS with dependencies.
See discussion over at ossc-db/pg_plan_advsr#7 (comment) .

[rjuju]

Thanks, as mentioned in the other thread that's already in my todo list, and I expect to work on this shortly!