OpenText Analytics Database 26.2.x – Working with spatial objects in tables

Data-Analysis: Defining table columns for spatial data

Mon, 01 Jan 0001 00:00:00 +0000

To define columns to contain GEOMETRY and GEOGRAPHY data, use this command:

=> CREATE TABLE [[db-name.]schema.]table-name (
   column-name GEOMETRY[(length)],
   column-name GEOGRAPHY[(length)]);

If you omit the length specification, the default column size is 1 MB. The maximum column size is 10 MB. The upper limit is not enforced, but the geospatial functions can only accept or return spatial data up to 10 MB.

You cannot modify the size or data type of a GEOMETRY or GEOGRAPHY column after creation. If the column size you created is not sufficient, create a new column with the desired size. Then copy the data from the old column, and drop the old column from the table.

You cannot import data to or export data from tables that contain spatial data from another database.

Important

A column width that is too large could impact performance. Use a column width that fits the data without being excessively large. See STV_MemSize.

Data-Analysis: Exporting spatial data from a table

Mon, 01 Jan 0001 00:00:00 +0000

You can export spatial data from a table in your database to a shapefile.

To export spatial data from a table to a shapefile:

As the superuser., set the shapefile export directory.

=> SELECT STV_SetExportShapefileDirectory(USING PARAMETERS path = '/home/geo/temp');
               STV_SetExportShapefileDirectory
------------------------------------------------------------
 SUCCESS. Set shapefile export directory: [/home/geo/temp]
(1 row)

Export your spatial data to a shapefile.

=> SELECT STV_Export2Shapefile(*
              USING PARAMETERS shapefile = 'visualizations/city-data.shp',
                            shape = 'Polygon') OVER() FROM spatial_data;
 Rows Exported |                          File Path
---------------+----------------------------------------------------------------
       185873 | v_geo-db_node0001: /home/geo/temp/visualizations/city-data.shp
(1 row)

The value asterisk (*) is the equivalent to listing all columns in the FROM clause.
You can specify sub-directories when exporting your shapefile.
Your shapefile must end with the file extension .shp.

Verify that three files now appear in the shapefile export directory.
```
$ ls
city-data.dbf  city-data.shp   city-data.shx
```

Data-Analysis: Identifying null spatial objects

Mon, 01 Jan 0001 00:00:00 +0000

You can identify null GEOMETRY and GEOGRAPHY objects using the IS NULL and IS NOT NULL constructs.

This example uses the following table, where the row with id=2 has a null value in the geog field.


=> SELECT id, ST_AsText(geom), ST_AsText(geog) FROM locations
   ORDER BY 1 ASC;
 id |           ST_AsText              |                 ST_AsText
----+----------------------------------+--------------------------------------
  1 | POINT (2 3)                      | POINT (-85 15)
  2 | POINT (4 5)                      |
  3 | POLYGON ((-1 2, 0 3, 1 2, -1 2)) | POLYGON ((-24 12, -15 23, -20 27, -24 12))
  4 | LINESTRING (-1 2, 1 5)           | LINESTRING (-42.74 23.98, -62.19 23.78)
(4 rows)

Identify all the rows that have a null geog value:

=> SELECT id, ST_AsText(geom), (ST_AsText(geog) IS NULL) FROM locations
   ORDER BY 1 ASC;
 id |            ST_AsText             | ?column?
----+----------------------------------+----------
  1 | POINT (2 3)                      | f
  2 | POINT (4 5)                      | t
  3 | POLYGON ((-1 2, 0 3, 1 2, -1 2)) | f
  4 | LINESTRING (-1 2, 1 5)           | f
(4 rows)

Identify the rows where the geog value is not null:

=> SELECT id, ST_AsText(geom), (ST_AsText(geog) IS NOT NULL) FROM locations
   ORDER BY 1 ASC;
 id |            st_astext             | ?column?
----+----------------------------------+----------
  1 | POINT (2 3)                      | t
  2 | POINT (4 5)                      | f
  3 | LINESTRING (-1 2, 1 5)           | t
  4 | POLYGON ((-1 2, 0 3, 1 2, -1 2)) | t
(4 rows)

Data-Analysis: Loading spatial data from shapefiles

Mon, 01 Jan 0001 00:00:00 +0000

OpenText™ Analytics Database provides the capability to load and parse spatial data that is stored in shapefiles. Shapefiles describe points, lines, and polygons. A shapefile is made up of three required files; all three files must be present and in the same directory to define the geometries:

.shp—Contains the geometry data.
.shx—Contains the positional index of the geometry.
.dbf—Contains the attributes for each geometry.

To load spatial data from a shapefile:

Use STV_ShpCreateTable to generate a CREATE TABLE statement.

=> SELECT STV_ShpCreateTable ( USING PARAMETERS file = '/home/geo/temp/shp-files/spatial_data.shp')
                               OVER() AS spatial_data;
           spatial_data
----------------------------------
 CREATE TABLE spatial_data(
   gid IDENTITY(64) PRIMARY KEY,
   uniq_id INT8,
   geom GEOMETRY(85)
);
(5 rows)

Create the table.

=> CREATE TABLE spatial_data(
   gid IDENTITY(64) PRIMARY KEY,
   uniq_id INT8,
   geom GEOMETRY(85));

Load the shapefile.

=> COPY spatial_data WITH SOURCE STV_ShpSource(file='/home/geo/temp/shp-files/spatial_data.shp')
    PARSER STV_ShpParser();
 Rows Loaded
-------------
          10
(1 row)

Supported shapefile shape types

The following table lists the shapefile shape types that OpenText™ Analytics Database supports.

Shapefile Shape Type	Supported
Null shape	Yes
Point	Yes
Polyline	Yes
Polygon	Yes
MultiPoint	Yes
PointZ	No
PolylineZ	No
PolygonZ	No
MultiPointZ	No
PointM	No
PolylineM	No
PolygonM	No
MultiPointM	No
MultiPatch	No

Data-Analysis: Loading spatial data into tables using COPY

Mon, 01 Jan 0001 00:00:00 +0000

You can load spatial data into a table in the database using a COPY statement. Perform the following steps:

Create a table.

=> CREATE TABLE spatial_data (id INTEGER, geom GEOMETRY(200));
CREATE TABLE

Create a text file named spatial.dat with the following data.

1|POINT(2 3)
2|LINESTRING(-1 2, 1 5)
3|POLYGON((-1 2, 0 3, 1 2, -1 2))

Use COPY to load the data into the table.
```
=> COPY spatial_data (id, gx FILLER LONG VARCHAR(605), geom AS ST_GeomFromText(gx)) FROM LOCAL 'spatial.dat';
 Rows Loaded
-------------
           3
(1 row)
```
The statement specifies a LONG VARCHAR(32000000) filler, which is the maximum size of WKT. You must specify a filler value large enough to hold the largest WKT you want to insert into the table.

Data-Analysis: Retrieving spatial data from a table as well-known text (WKT)

Mon, 01 Jan 0001 00:00:00 +0000

GEOMETRY and GEOGRAPHY data is stored in database tables as LONG VARBINARY, which isn't human readable. You can use ST_AsText to return the spatial data as Well-Known Text (WKT).

To return spatial data as WKT:

=> SELECT id, ST_AsText(geom) AS WKT FROM spatial_data;
 id |               WKT
----+----------------------------------
  1 | POINT (2 3)
  2 | LINESTRING (-1 2, 1 5)
  3 | POLYGON ((-1 2, 0 3, 1 2, -1 2))
(3 rows)

Data-Analysis: Working with GeoHash data

Mon, 01 Jan 0001 00:00:00 +0000

OpenText™ Analytics Database supports GeoHashes. A GeoHash is a geocoding system for hierarchically encoding increasingly granular spatial references. Each additional character in a GeoHash drills down to a smaller section of a map.

You can use the database to generate spatial data from GeoHashes and GeoHashes from spatial data. The following functions are supported for use with GeoHashes:

ST_GeoHash - Returns a GeoHash in the shape of the specified geometry.
ST_GeomFromGeoHash - Returns a polygon in the shape of the specified GeoHash.
ST_PointFromGeoHash - Returns the center point of the specified GeoHash.

For example, to generate a full precision and partial precision GeoHash from a single point.

=> SELECT ST_GeoHash(ST_GeographyFromText('POINT(3.14 -1.34)')), LENGTH(ST_GeoHash(ST_GeographyFromText('POINT(3.14 -1.34)'))),
                     ST_GeoHash(ST_GeographyFromText('POINT(3.14 -1.34)') USING PARAMETERS numchars=5) partial_hash;
      ST_GeoHash      | LENGTH | partial_hash
----------------------+--------+--------------
 kpf0rkn3zmcswks75010 |     20 | kpf0r
(1 row)

This example shows how to generate a GeoHash from a multipoint point object. The returned polygon is a geometry object of the smallest tile that encloses that GeoHash.

=> SELECT ST_AsText(ST_GeomFromGeoHash(ST_GeoHash(ST_GeomFromText('MULTIPOINT(0 0, 0.0002 0.0001)')))) AS region_1,
                    ST_AsText(ST_GeomFromGeoHash(ST_GeoHash(ST_GeomFromText('MULTIPOINT(0.0001 0.0001, 0.0003 0.0002)')))) AS region_2;
 -[ RECORD 1 ]---------------------------------------------------------------------------------------------
    region_1 | POLYGON ((0 0, 0.00137329101562 0, 0.00137329101562 0.00137329101562, 0 0.00137329101562, 0 0))
    region_2 | POLYGON ((0 0, 0.010986328125 0, 0.010986328125 0.0054931640625, 0 0.0054931640625, 0 0))

Data-Analysis: Spatial joins with ST_Intersects and STV_Intersect

Mon, 01 Jan 0001 00:00:00 +0000

Spatial joins allow you to identify spatial relationships between two sets of spatial data. For example, you can use spatial joins to:

Calculate the density of mobile calls in various regions to determine the location of a new cell phone tower.
Identify homes that fall within the impact zone of a hurricane.
Calculate the number of users who live within a certain ZIP code.
Calculate the number of customers in a retail store at any given time.