[Scipy-svn] r4191 - trunk/scipy/cluster
Sun Apr 27 08:06:17 CDT 2008
Date: 2008-04-27 08:06:16 -0500 (Sun, 27 Apr 2008)
New Revision: 4191
Tightened language in vq's module summary docstring.
--- trunk/scipy/cluster/vq.py 2008-04-27 12:55:26 UTC (rev 4190)
+++ trunk/scipy/cluster/vq.py 2008-04-27 13:06:16 UTC (rev 4191)
@@ -51,13 +51,19 @@
is the code corresponding to the code_book[i] centroid.
- Normalize a group of observations so each feature has unit variance.
+ Normalize a group of observations so each feature has unit
- Calculate code book membership of obs.
+ Calculate code book membership of a set of observation
- Train a codebook for mimimum distortion using the k-means algorithm.
+ Clusters a set of observation vectors. Learns centroids with
+ the k-means algorithm, trying to minimize distortion. A code
+ book is generated that can be used to quantize vectors.
- Similar to kmeans, but with several initialization methods.
+ A different implementation of k-means with more methods for
+ initializing centroids. Uses maximum number of iterations as
+ opposed to a distortion threshold as its stopping criterion.
__docformat__ = 'restructuredtext'
@@ -580,10 +586,10 @@
centroids to generate. If minit initialization string is
'matrix', or if a ndarray is given instead, it is
interpreted as initial cluster to use instead.
- niter : int
- Number of iterations of k-means to run. Note that this
- differs in meaning from the iters parameter to the kmeans
+ iter : int
+ Number of iterations of the k-means algrithm to run. Note
+ that this differs in meaning from the iters parameter to
+ the kmeans function.
thresh : float
(not used yet).
minit : string
More information about the Scipy-svn