Re: Why Varying Arrasy and Nested Tables?

Interesting. But not sure about the "efficient" aspect of it. Can you clarify why it is more efficient to have nested table? In your case, you still need 20 rows in that nested table, no? Did you compare the pros and cons of creating a separate "dimension" table for date_time, instead of a nested table?

Storing data in one consideration. Retrieving it is another. I figure you will want to query against the DO data, maybe perform some sort of aggregate on it. Seems to me the aggregates are going to give you some trouble.

• Jusung Yang

"Bryan & Sarah Baker" <bakers65_at_cox.net> wrote in message news:<tLAq9.42271$XF.24866_at_news1.central.cox.net>...
> We use nested tables in our database design and it has made things a lot
> easier for us since we have done theis.
>
> Say you have data that you collect over time for comparison or computation.
>
> Table = time_series_values (columns = date_time, value, location_code)
>
> Now lets say you sample Dissolved Oxygen and temperature in a lake at a
> specific point (Lat Long = location_code). In your sampling, you do those
> two measurements at different depths. After you are done you have
>
> Depth, DO, Temp
> 1, 2, 60
> 2, 2.5,59
> etc
>
>
> Well if you have value in the time_series_values tables setup as a nested
> table of (depth, Dissolved Oxygen, temp), you have have a more efficient
> normalized table then if your time_series table was (date_time, depth, DO,
> Temp, location code = table without nested tables) Why? Well, if your
> sampling had 20 measurements you would have 1 row with the nested table or
> 20 rows with out nested tables.
>
> We like nested tables over nested arrays since you set your array size and
> you can exceed that without recreating the table. Nested tables are
> variable in length. You can have as many or as few as you want/need.
>
> Bryan Baker
>
>
>
Received on Mon Oct 14 2002 - 21:12:55 CEST