> I wrote, successfully, a "date/time" query based on the tick data some
> time ago. I suspected, through my reading of the documentation, that the
> "narrowing of the fields--" so to speak--would be more complex than a
> simple WHERE statement.
Glad you were able to get it working.
I would suggest though, if you have the flexibility to change your
database structure, you might be better off with a single table like:
date time symbol open high low close volume
| 2009-01-15 | 16:49:00 | goog | 299.32 | 299.32 | 299.32 | 299.32 | 550 |
| 2009-01-15 | 16:51:00 | aapl | 299.93 | 299.93 | 299.93 | 299.93 | 500 |
| 2009-01-15 | 16:55:00 | msft | 299.4 | 299.4 | 299.39 | 299.39 | 269 |
(obviously those numbers are wrong in this example)
This way you don't need to redundantly store your dates and times and
your cross-symbol queries become much simpler. Any query you had
before which went against one of those individual tables can be
adapted with an additional where clause filter.
Good luck,
tim
_______________________________________________
PLUG mailing list
[email protected]
http://lists.pdxlinux.org/mailman/listinfo/plug
