[mapserver-commits] r11451 - trunk/docs/en/mapfile

[svn at osgeo.org]

Author: tamas
Date: 2011-04-03 10:06:18 -0700 (Sun, 03 Apr 2011)
New Revision: 11451

Modified:
   trunk/docs/en/mapfile/cluster.txt
Log:
Add doc for the aggregate functions (#3700)

Modified: trunk/docs/en/mapfile/cluster.txt
===================================================================
--- trunk/docs/en/mapfile/cluster.txt	2011-04-03 16:48:56 UTC (rev 11450)
+++ trunk/docs/en/mapfile/cluster.txt	2011-04-03 17:06:18 UTC (rev 11451)
@@ -8,7 +8,7 @@
 the layer into single (aggregated) features based on their relative positions.
 Supported only for POINT layers.
 
-The clustered layer itself provides the following aggregate attributes: 
+The clustered layer itself provides the following aggregated attributes: 
 
 1) Cluster:FeatureCount - count of the features in the clustered shape 
 2) Cluster:Group - The group value of the cluster (to which the group expression is evaluated)
@@ -18,8 +18,11 @@
 The ITEMS processing option can be used to specify a subset of the attributes from the original 
 layer in the query operations according to the user's preference.
 
-If we retrieve the original attributes then the cluster layer provider will provide only those values which
-are equal for each shapes contained by the cluster. The other values are set to "Cluster:Empty".
+We can use simple aggregate functions (Min, Max, Sum, Count) to specify how the clustered attribute should be
+calculated from the original attributes. The aggregate function should be specified as a prefix separated by ':' in
+the attibute definition, like: [Max:itemname]. If we don't specify aggregate functions for the source
+layer attributes, then the actual value of the cluster attribute will be non-deterministic if the cluster contains
+multiple shapes with different values. The Count aggregate function in fact provides the same value as Cluster:FeatureCount.
   
 .. seealso::       
     :ref:`rfc69`