Changes between Version 4 and Version 5 of DevClusteringFunctions


Ignore:
Timestamp:
02/02/15 12:41:18 (10 years ago)
Author:
Mike Taves
Comment:

fix formatting

Legend:

Unmodified
Added
Removed
Modified
  • DevClusteringFunctions

    v4 v5  
    1 Proposal for two clustering functions:
     1= Proposal for clustering functions =
     2
     3== `geometry[] ST_ClusterIntersecting(geometry geom)` ==
     4
     5Aggregate function returning an array of `GeometryCollection`s representing the connected components of a set of geometries.
     6
     7 - accepts `[Multi]Point`, `[Multi]LineString`, `[Multi]Polygon` geometries of any type that can be converted into GEOS (I can't think of a situation where `[Multi]Point` would be useful, but that doesn't mean there isn't one...)
     8 - return a geometry array (my current implementation returns a `GeometryCollection`, but the recursive semantics of `ST_Dump` then undo all of the hard work)
     9Example: if run on a table containing all of the `LineString`s in the image below, would return an array with two `MultiLineString` geometries (red and blue)
     10
     11[[Image(http://i.stack.imgur.com/WNlxX.png)]]
    212
    313----
    414
    5 {{{geometry[] ST_ClusterIntersecting(geometry geom)}}}
     15== `geometry[] ST_ClusterWithin(geometry geom, double precision distance)` ==
    616
    7 Aggregate function returning an array of GeometryCollections representing the connected components of a set of geometries.
     17Aggregate function returning an array of `GeometryCollection`s?/`MultiPoint`s?, where any component is reachable from any other component with jump of no more than the specified distance.
    818
    9 - accepts Point/MultiPoint, LineString, MultiLineString, Polygon, MultiPolygon geometries of any type that can be converted into GEOS (I can't think of a situation where Point/MultiPoint would be useful, but that doesn't mean there isn't one...)[[br]]
    10 - return a geometry array (my current implementation returns a GeometryCollection, but the recursive semantics of ST_Dump then undo all of the hard work)[[br]]
    11 Example: if run on a table containing all of the LineStrings in the image below, would return an array with two MultiLineString geometries (red and blue)
     19 - like `ST_ClusterIntersecting`, but uses a distance threshold rather than intersection when determining if two geometries should be included in the same component.  Could have an implementation very similar to `ST_ClusterIntersecting`, or could be restricted to points and maybe have a more efficient implementation.
     20 - differs from k-means in that a distance is provided, not a number of clusters
    1221
    13 [[http://i.stack.imgur.com/WNlxX.png]]
     22Example:  In the picture below, an array of five `MultiPoint`s would be returned (color-coded).  The threshold distance in this case was more than the orange line but less than the pink line.
    1423
    15 
    16 ----
    17 
    18 
    19 {{{geometry[] ST_ClusterWithin(geometry geom, double precision distance)}}}
    20 
    21 Aggregate function returning an array of GeometryCollections?/MultiPoints?, where any component is reachable from any other component with jump of no more than the specified distance.
    22 
    23 - like ST_ClusterIntersecting, but uses a distance threshold rather than intersection when determining if two geometries should be included in the same component.  Could have an implementation very similar to ST_ClusterIntersecting, or could be restricted to points and maybe have a more efficient implementation.[[br]]
    24 - differs from kmeans in that a distance is provided, not a number of clusters[[br]]
    25 
    26 Example:  In the picture below, an array of five MultiPoints would be returned (color-coded).  The threshold distance in this case was more than the orange line but less than the pink line.
    27 
    28 [[http://ibin.co/1oH1ApWCoW8L]]
     24[[Image(http://ibin.co/1oH1ApWCoW8L)]]