Re: SQL: Working with huge tables of chronological data

"John" <acide.ascorbique_at_gmail.com> wrote in news:1175783243.167337.195580 @n59g2000hsh.googlegroups.com:

> Hi,
>
> I'm trying to figure out an efficient way to search for the non
> existence of events in chronological data with SQL. The goal (detailed
> below) seems pretty simple but so far it looks like it's tricky to do
> with Oracle. Here's my problem:
>
> I'm working with 2 simple but huge tables each recording a different
> kind of event associated with a timestamp. For instance:
>
> Table A
> (03:50pm, A1)
> (03:55pm, A2)
> (03:58pm, A3)
>
> Table B
> (03:51pm, B1)
> (04:00pm, B2)
>
> I'm looking for all the chronological sequences (Ax, Ay) where no B
> event is present between Ax and Ay. In this example, the result would

I do not understand the logic that would lead the answer above; since none of the timestamps in Table A match any timestamp in Table B.

> be (A2, A3).
>
> I've been searching actively for an efficient solution for this
> problem and I couldn't find any fast enough. Do you have any idea?

How fast is fast enough?
How do we know what you tried & deemed unacceptable? Received on Thu Apr 05 2007 - 09:51:28 CDT