Basic Spatial Query

Question

Hi guys,
I am currently developing a java applet to display telecommunication sites taken from an Oracle 8.1.6 DB. The company I work for also happens to have the spatial cartridge which I intend to use to overlay a country map over my plottings. I have imported the map data into a table, and I can view the GEOLOC information as a huge stream of longs and lats via a simple SELECT query in SQLPLUS - so the data has been imported correctly.
I have tried to digest huge articles on querying spatial information - however, unfortunately being an Oracle novice, it has left me quite dazzled.
Basically all I want to do is take from my spatial table a list of long and lat coordinates (in double format) in a specific physical area, so that I can display them from there.
I'd appreciate any help that anyone could give.
null

Answer 1

Hi David,
There are a few things you will have to do in order to get this to work quickly.
1) You will have to insert data into the user_sdo_geom_metadata field so Oracle Spatial can know some information about the data you've created (table name, spatial column name, information about upper and lower bounds as well as tolerance for each dimension.
2) You'll need to create a spatial index on the data.
3) You'll need to analyze the spatial index table.
4) Then you can query the data. You'll have to decide whether you need exact answers (using the sdo_relate operator) or approximate answers based only on the index (sdo_filter operator).
There is too much information required to put it all here, but I'll include a few samples.
to insert data into user_sdo_geom_metadata:
insert into user_sdo_geom_metadata values ('TABLE_NAME','GEOMETRY_COLUMN_NAME',
mdsys.sdo_dim_array(mdsys.sdo_dim_element('X',-180.0,180,0.00000005),
mdsys.sdo_dim_element('Y',-90,90,0.00000005)),NULL);
commit;
In the above example, TABLE_NAME is the name of the table with your geometry column, then the geometry column name, then the dimarray has attributes in it associated with each of two dimensions (two dimensional data is stored here), the first dimension has a lower bound of -180, and upper bound of 180, and a tolerance or 0.00000005 (the decimal precision of your data, with a 5 appended), the second dimension has the info supplied, and the last null means there is no coordinate system stored with the data. I'd suggest if you are a newbie to leave it as NULL, then ease into coordinate systems once you get more experience.
2) create the index
create index table_name_sidx
on table_name(geometry_column_name)
indextype is mdsys.spatial_index
parameters('sdo_level=x');
where x is usually a number between 6 and 14 based on the type of data you have. There is a lot to say about this, but I can't say it here.
3) analyze the spatial index table:
from sql*plus, run the following statements:
set head off
spool analyze_it.sql
select ' analyze table ' | | sdo_index_table | | ' estimate statistics sample 32 percent;'
from user_sdo_index_metadata order by sdo_index_table;
spool off;
@analyze_it.sql
4) execute a query based on an area of interest:
this statement will return all geometries that have any interaction with a query window specified by a rectangle with bounds
2,2, 10,2, 10,10, 2,10, 2,2
SELECT geometry_column_name
FROM table_name
where sdo_relate (geometry_column_name,
mdsys.sdo_geometry(2003,NULL, NULL
mdsys.sdo_elem_info_array(1,1003,1),
mdsys.sdo_ordinate_array (2,2, 10,2, 10,10, 2,10, 2,2)),
'mask=anyinteract querytype=window'
= 'TRUE';
You will have to dig more into each of these subject to understand them.
Hope this helps,
dan
null