First up I just wanted to reply as it has been a couple of days. We are very much open to contributions and would far rather you work as part of the community then maintain your own fork.
On the specifics I do not have much of an opinion as I do not use the point stacker often :) -- Jody Garnett On Mon, 15 Jun 2020 at 03:10, Sabrina Arnold <sarn...@codematix.de> wrote: > Dear community, > > regarding one of our projects we utilized your open-source *PointStacker* > and adjusted the source code to our needs. > First of all, thanks a lot for providing this highly helpful code as > open-source code! If you are interested, we would > like to contribute to the development of the code with our extensions. We > added input parameters, and > output attributes that might be requested for other use cases as well. If > you are interested in these code extensions, I > would be happy to create a Merge Request. I am refering to > /org/geotools/process/vector/PointStackerProcess.java > <https://github.com/geotools/geotools/blob/master/modules/unsupported/process-feature/src/main/java/org/geotools/process/vector/PointStackerProcess.java> > . > Here you can see our customized PointStackerProcess as forked repository > <https://github.com/SaCodematix/geotools/blob/master/modules/unsupported/process-feature/src/main/java/org/geotools/process/vector/PointStackerProcess.java> > as well as the additional unit tests > <https://github.com/SaCodematix/geotools/blob/master/modules/unsupported/process-feature/src/test/java/org/geotools/process/vector/PointStackerProcessTest.java> > (see "CM_POINSTACKER" comments). > > In the following you can find more detailed information about our code > extensions: > > (Definition: The data set consists of features that are stacked to > cluster. A cluster point then is representing the assigned features. A > StackedPoint contains a feature that is stacked to a cluster. Original > attributes are column names of the data set connected through the GeoServer > datastore. Single points are features that are not assigned to a cluster > with other features, thus, count equals 1.) > > In addition to the original source code and the clustering by grid, our > adjustments allow GeoServer users to cluster by an original attribute, > such as state or district (depending on the given data set). The already > existing location options 'nearest' and 'weighted' of how to position the > cluster point are expanded by 'average' (average of all coordinates of > the clustered features) and 'extent' (center of the minimum and maximum > coordinates in x and y direction). For this, an additional input > parameter *clusterBasis* is added. Depending on the data and use case, these > locating approaches result in more suitable cluster point locations on the > final map visualization. > > The output is expanded by additional returning values. The fids ( > *listStackedPointsIDs*) and coordinates (*listStackedPtsCoos*) of the > features belonging to the cluster are listed, as well as all available > properties (attributes & values) for single points ( > *singlePointOrigAttributes*) which might be required as FeatureInfo on > the final map. > > If the user wants to sort the resulting data collection (*sortedByField > *by overwriting default values of result.SortBy()) by a specified original > attribute field, for instance grades from 1 to 6, three additional > parameters are introduced: *sortField* (e.g. 'grade'), > *sortBy* (can be either 'DESCENDING' or 'ASCENDING' (default)), and > *sortValueStackedPt* to set a representative value for cluster points > that otherwise would contain a list of grades, instead of a single number that > can be sorted. We required this sorting method to influence the rendered > z-order of several PointSymbolizers in the style .sld file, since the > available VendorOption SortBy > <https://docs.geoserver.org/latest/en/user/styling/sld/extensions/z-order/syntax.html> > option had no influence on the z-order of the PointSymbolizers in our > single FeatureTypeStyle since a transformation (PointStacker) is included. > > With these extensions, the user also has the possibility to add original > attributes to the output result (*originalAttributes)*. For instance, if > the following list is given in the style as list of original attributes, > > <ogc:Function name="parameter"> > <ogc:Literal>originalAttributes</ogc:Literal> > <ogc:Literal>*attrA,attrB*,*attrC*</ogc:Literal> > </ogc:Function> > > the *extended output* is a FeatureCollection containing the following > attributes: > > - fid - the fid of the point representing the cluster > - count - the total number of points in the cluster > - countunique - the number of unique point locations in the cluster > - envBBOX - bounding box coordinates > - *listStackedPointsIDs* - list of all IDs of the stacked features in > the cluster > - *listStackedPtsCoos* - list of all coordinates of the stacked > features in the cluster > - *singlePointOrigAttributes* - only for single points, all of its > data properties (attributes & values) > - *sortedByField* contains the original attribute values of the > sortField, or the representative sortValueStackedPt for cluster points, or > nothing, if no sortField is indicated > - *attrA* - list of values for the first attribute of the input > parameter originalAttributes > - *attrB* - list of values for the second attribute of the input > parameter originalAttributes > - *attrC* - list of values for the third attribute of the input > parameter originalAttributes, and so on. > > To approach this output, the following input parameters are added: > > @DescribeParameter( > name = "*clusterBasis*", > description = "Indicate basis for clustering methods: 'grid' > (default), or an original " + > "attribute such as state or district (data-dependent).", > defaultValue = "grid" > ) > @DescribeParameter( > name = "*originalAttributes*", > description = "List of original attributes that are requested to > be returned. If these " + > "indicated attributes are no actual original attributes > of data, they will be ignored. " + > "Required as comma-separated list of format, e.g: a,b,c. > All in lower case!", > min = 0 > ) > String argOrigAttributes, > @DescribeParameter( > name = "*sortField*", > description = "Attribute field by which resulting collection > (optionally) can be sorted.", > min = 0 > ) > String argSortField, > @DescribeParameter( > name = "*sortBy*", > description = "If collection is desired to be sorted through a > specified 'sortField', one can " + > "indicate the method to sort, either 'DESCENDING' or > 'ASCENDING' (default) .", > min = 0 > ) > String argSortOrder, > @DescribeParameter( > name = "*sortValueStackedPt*", > description = "If a collection is desired to be sorted vy a > specified column 'sortField', here" + > "you can indicate the value that will be used for > clustered points as sort value since only" + > "single points will keep the original value of the > 'sortField' while clustered points " + > "gather the values of the 'sortField's to an array which > will be replaced by this value " + > "instead, such as 0 (default).", > min = 0 > ) > String argSortValueStackedPt > > > Please let me know, if you have any questions or unclarity. I am looking > forward to any responses, discussions and more information about a possible > Merge Request. > > > -- > > All the best > > Sabrina Arnold > [image: logo] > Felsbachstr. 5/7 > D-07745 Jena > Tel: +49 (3641) 30 38-12 > Fax: +49 (3641) 30 38-20 > Internet: www.codematix.de > e-mail: sarn...@codematix.de > > ------------------------------------------------------------------------------------------------------------ > Handelsregister: Amtsgericht Jena / HRB 10575 / Geschäftsführung: Carsten > Busch > > Diese E-Mail könnte vertrauliche und / oder rechtlich geschützte > Informationen enthalten. > Wenn Sie nicht der richtige Adressat sind oder diese E-Mail irrtümlich > erhalten haben, > informieren Sie bitte sofort den Absender und vernichten Sie diese Mail. > Das unerlaubte Kopieren sowie die unbefugte Weitergabe dieser Mail sind > nicht gestattet. > > This e-mail may contain confidential and/or privileged information. > If you are not the intended recipient (or have received this e-mail in > error) > please notify the sender immediately and destroy this e-mail. > Any unauthorised copying, disclosure or distribution of the material in > this e-mail is strictly forbidden. > > Please consider the environment - only print this e-mail if absolutely > necessary! > _______________________________________________ > GeoTools-Devel mailing list > GeoTools-Devel@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/geotools-devel >
_______________________________________________ GeoTools-Devel mailing list GeoTools-Devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/geotools-devel
