Added more info on the data input variable of kmeans
authorStevenPuttemans <steven.puttemans@gmail.com>
Tue, 13 May 2014 12:35:08 +0000 (14:35 +0200)
committerStevenPuttemans <steven.puttemans@gmail.com>
Thu, 22 May 2014 10:56:44 +0000 (12:56 +0200)
modules/core/doc/clustering.rst

index 0f9fa6c..49fd1e2 100644 (file)
@@ -17,7 +17,15 @@ Finds centers of clusters and groups input samples around the clusters.
 
     :param samples: Floating-point matrix of input samples, one row per sample.
 
-    :param data: Data for clustering.
+    :param data: Data for clustering. An array of N-Dimensional points with float coordinates is needed. Examples of this array can be:
+
+           * ``Mat points(count, 2, CV_32F);``
+
+           * ``Mat points(count, 1, CV_32FC2);``
+
+           * ``Mat points(1, count, CV_32FC2);``
+
+           * ``std::vector<cv::Point2f> points(sampleCount);``
 
     :param cluster_count: Number of clusters to split the set by.