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`(*)](https://help.sap.com/viewer/bc9e455fe75541b8a248b4c09b086cf5/LATEST/en-US/57758b2af95346db9a478a53ec2c4ccb.html).

```SQL
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`(*)](https://help.sap.com/viewer/bc9e455fe75541b8a248b4c09b086cf5/LATEST/en-US/010c53e227a94966bb009d52d9ec47a2.html).

```SQL
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;

</div><div class="step-validation-mount"
      data-step="1"></div>
 <div class="step-actions">
   <button class="fd-button fd-button--emphasized" data-action="mark-done" data-step="1">Done</button>
 </div>
</div>
</div>






<div id="step-2" class="tutorial-step" data-step="2">
<div class="step-header" data-action="toggle-step">
 <span class="step-check-circle"></span>
 <div class="step-header-text">
   <span class="step-label">Step 2</span>
   <span class="step-title-text">Add flags for all nodes in a circle</span>
 </div>
 <span class="step-toggle-icon">+</span>
</div>
<div class="step-body" hidden>
 <hr class="step-divider" />
 <div class="step-content">

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
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
SELECT SHAPE FROM LONDON_VERTICES WHERE IN_SCOPE = 1;

![relevant nodes](https://raw.githubusercontent.com/sap-tutorials/Tutorials/master/tutorials/hana-cloud-smart-multi-model-4/ss-02-relevant-nodes.png)

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.