[Qgis-user] Draw Ellipse/Polygons for areas of maximum concentration

[Bernd Vogelgesang]

As you mention degree as unit for the distance, I assume that you are
working with unprojected data. 1 degree is a huge distance, so
consequently, all points will fall into this area and form one cluster.

You'll have to project your data and find some meaningful value for the
distance to get different clusters.

On 30.04.21 02:02, krishna Ayyala wrote:
> Nyall,
> Thanks for your reply. I have used the "dbscan clustering". In this, the
> default values for "Minimum Cluster Size" is 5 and Max distance between
> cluster points is 1 degrees. I ran with the same default values. I am
> getting output with the layer name "Clusters". This has generated a new
> field "Cluster_ID". All the points have same Cluster_ID i.e. 1. This has
> also generated new field called "Cluster_Size". All the points have same
> cluster size i.e. 1699. Can you please help me on how to get 4 different
> cluster ID's.
>
> Regards.
>
> On Thu, Apr 29, 2021 at 4:29 PM Nyall Dawson <nyall.dawson at gmail.com> wrote:
>
>>
>> On Fri, 30 Apr 2021 at 08:25, krishna Ayyala <ayyalakrishna at gmail.com>
>> wrote:
>>
>>> Chris,
>>> Thanks for the reply. Yes, I did run k-means clustering for 4 clusters.
>>> It is creating a new layer called "clusters". This layer has a field called
>>> Cluster_id, ranging from 1 to 4. But, this method is considering all the
>>> points within the circle. I am looking for the points outside the ellipses
>>> to be omitted (should not be considered)
>>>
>> In this case "dbscan clustering" is more appropriate.
>>
>>
>>> . Also K-means clustering is not generating any polygons/ellipses. We
>>> have to identify a cluster based on the Cluster_ID. I am curious if there
>>> is any tool within Qgis that can produce results similar to the circle with
>>> ellipses?
>>>
>> What you could do is dissolve the points based on the cluster_id field,
>> and then generate convex (or concave) hulls enclosing each set of points.
>> You won't get ellipses, but you'll get polygons describing the boundaries.
>> (And it would be relatively straightforward to wrap up these steps into a
>> single graphical model so that you have one tool which gives the desired
>> output!).
>>
>> Nyall
>>
>>
>>
>>
>>> Regards.
>>>
>>> On Thu, Apr 29, 2021 at 3:56 PM chris hermansen <clhermansen at gmail.com>
>>> wrote:
>>>
>>>> Krishna and list,
>>>>
>>>> On Thu, Apr 29, 2021 at 2:51 PM krishna Ayyala <ayyalakrishna at gmail.com>
>>>> wrote:
>>>>
>>>>> Hello,
>>>>> I have a circle in which there are randomly distributed points as
>>>>> below. Each point is a customer.
>>>>>
>>>>> [image: image.png]
>>>>>
>>>>>
>>>>> Is there a tool within QGIS that can automatically generate four
>>>>> polygons or four ellipses such as below. These polygons are the areas of
>>>>> maximum concentration of the customers?
>>>>>
>>>>> [image: image.png]
>>>>>
>>>>> You could try k-means clustering
>>>>
>>>> https://docs.qgis.org/3.4/en/docs/user_manual/processing_algs/qgis/vectoranalysis.html#k-means-clustering
>>>>
>>>> --
>>>> Chris Hermansen · clhermansen "at" gmail "dot" com
>>>>
>>>> C'est ma façon de parler.
>>>>
>>> _______________________________________________
>>> Qgis-user mailing list
>>> Qgis-user at lists.osgeo.org
>>> List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
>>> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user
>>>
>
> _______________________________________________
> Qgis-user mailing list
> Qgis-user at lists.osgeo.org
> List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/qgis-user/attachments/20210430/7f54182c/attachment.html>