Identify Relevant Networks in Spatial Data

First, you need to construct a circle to identify the relevant area for the transportation network. The circle should have a minimal size and should contain origin and destination with a buffer of 500 meters around it. To illustrate how the query is being built we will use step-by-step queries with sub-SELECT statements.

Note, that generally, the center point of the smallest circle, which includes two points, is the center point of their direct connection line.

You can see in this image the circle you will create:

1. First, select the two points from the previous tutorials using this statement:

   SQL

   Copy

   SELECT
   ST_GeomFromText('POINT (706327.107445 5710259.94449)', 32630) AS START_PT,
   SHAPE AS TARGET_PT
   FROM LONDON_POI lp
   WHERE "osmid" = 6274057185;
   

2. To create a line geometry, which is connecting both points, you can then use the function ST_MakeLine(*).

   SQL

   Copy

   SELECT ST_MakeLine(START_PT, TARGET_PT) AS CONN_LINE
   FROM
   (
   -- previous statement
   );
   

3. To retrieve an arbitrary point on this line, we need to use function ST_LineInterpolatePoint(*), which takes a fraction of the line as argument. For retrieving the center point of the line, we pass the value 0.5.

   SQL

   Copy

   SELECT CONN_LINE.ST_LineInterpolatePoint(0.5) AS CENTER_PT
   FROM
   (
   -- previous statement
   );
   

4. As a final step, we would like to draw a circle with radius distance (start, target)/2 + 500 around CENTER_PT. The respective function, which takes the radius as an input, is called ST_Buffer(*).

   SQL

   Copy

   SELECT CENTER_PT.ST_Buffer(4835) AS AREA
   FROM
   (
   -- previous statement
   );
   

5. The above steps can be combined into a single select using method chaining.

   SQL

   Copy

   SELECT
   ST_MakeLine(
       ST_GeomFromText('POINT (706327.107445 5710259.94449)', 32630),
       SHAPE
   )
   .ST_LineInterpolatePoint(0.5)
   .ST_Buffer(5000) AS AREA
   FROM LONDON_POI
   WHERE "osmid" = 6274057185;
   

Next, you need to flag all nodes of the transportation network located in the circle, by setting in_scope= true.

1. First, we need to enhance the existing vertex table by column in_scope.

   SQL

   Copy

   ALTER TABLE LONDON_VERTICES ADD (IN_SCOPE INTEGER);
   

2. Second, we fill the column with 0 or 1 based on the intersection with the identified area. We can use function ST_Intersects(*) to determine if two geometries intersect or not.

3. Instead of using an UPDATE statement, we will use MERGE INTO, which allows a more complex update logic.

   SQL

   Copy

   MERGE INTO LONDON_VERTICES lv
   USING
   (
   -- previous statement begin --
   SELECT
   	ST_MakeLine(
       	ST_GeomFromText('POINT (706327.107445 5710259.94449)', 32630),
       	SHAPE
   	)
   	.ST_LineInterpolatePoint(0.5)
   	.ST_Buffer(5000) AS AREA
   FROM LONDON_POI
   WHERE "osmid" = 6274057185
   	-- previous statement end --
   ) circle ON 1=1
   WHEN MATCHED THEN UPDATE SET lv.IN_SCOPE = CIRCLE.AREA.ST_Intersects(SHAPE);
   

4. You can confirm that you properly selected the nodes by using your preferred tool for visualizing the result set with this query:

   SQL

   Copy

   SELECT SHAPE FROM LONDON_VERTICES WHERE IN_SCOPE = 1;
   

   

By creating a circle in your relevant area and flagging nodes in this circle, you have successfully identified a sub-network for transportation, which you will consider for finding your path through London in the following exercises.

Learn in the next tutorial how to check if this area is suitable for riding a bike by visualizing edges and using Voronoi cells.