I need a 4D (may be 5D or 6D) data space
Date: Wed, 09 Jan 2002 11:29:04 +0100
Message-ID: <3C3C1B70.1C35E5A_at_Hvidberg.net>
Newbee - Please bare...
I need to establish a database. I know what I want, but not how to get there.
I need a database to hold a number of parameters that I will access from
another program. The other program is a GIS (Geographical Information
System), but I don't think that is very important for my question.
I have a visual idea about the database that I need.
Considering a ordinary table as a two dimensional (2D) data space, with
the two dimensions represented by rows and cols, respectively. I need a
4D (may be 5D or 6D) data space. I should be able to hold a small number
of information for each of a limited number of situations in a forest
environment. The situation in the forest is defined by four parameters
each representing the four dimensions. A "soil type" (dimension 1), a
Getting information out of the database will (as I see it) require
looking up a location in a 4D data space. I know the soil type, tree
species, the management and precipitation. I need the data base to
return a number of values (less than ten floats) which are associated
with that given forest situation. I can think of this also as a function
returning a block of values e.g. F(soil, species, management,
precipitation)=Result. Where Result are a array or some other data
structure holding the relevant values.
The 4D data space are not even that big. We expect to be using 7
soiltypes, 5 tree species, 3 managements and 9 precipitation intervals.
This gives less than 1000 possible locations in the 4D data space.
The tricky part
Q: Is this possible - How do I do - What data structure is optimal?
"tree species" (dimension 2), a "management type" (dimension 2) and a
"precipitation value" (dimension 2).
There is a catch, of cause, why I don't implement this as a simple .CSV
file.
1) I need this to be accessible, in an easy way, from the GIS system.
Specifically, the programming language Avenue, from ESRI.
This in terms mean that a ODBC and SQL enabled access would be
preferable.
2) We need SPEED. To meet acceptable reply times for the system as such
we would like close to 10.000 records / second, out of the data base.
All comments and suggestions are welcomed...
Best Regards
Martin Hvidberg (mhv_at_fsl.dk)
PS. We have access to ms-access, a ms-SQL-server and a MySQL server. Received on Wed Jan 09 2002 - 11:29:04 CET