Changeset 21476 for trunk/psModules/src/imcombine/pmStack.h

Timestamp:

Feb 13, 2009, 1:52:14 PM (17 years ago)

Author:

Paul Price

Message:

Using covariance factor to set the correct variance level. Median does not do a good job of selecting the middle of the distribution when there's a variety of input weights (esp. when one image has dominant weight), so now using an 'olympic weighted mean' to provide the mean value for rejection. Tried a 'weighted median', but it doesn't do as good a job.

File:

• trunk/psModules/src/imcombine/pmStack.h (modified) (2 diffs)

trunk/psModules/src/imcombine/pmStack.h

@@ -8 +8 @@
  * @author GLG, MHPCC
  *
- * @version $Revision: 1.10 $ $Name: not supported by cvs2svn $
- * @date $Date: 2009-01-27 06:39:38 $
+ * @version $Revision: 1.11 $ $Name: not supported by cvs2svn $
+ * @date $Date: 2009-02-13 23:52:14 $
  *
  * Copyright 2004-2007 Institute for Astronomy, University of Hawaii
@@ -49 +49 @@
                     float rej,          ///< Rejection limit (standard deviations)
                     float sys,          ///< Relative systematic error
+                    float discard,      ///< Fraction of values to discard for Olympic weighted mean
                     bool entire,        ///< Combine entire image even if rejection lists provided?
                     bool useVariance,   ///< Use variance values for rejection?