Back to the main page.
Bug 2849 - volumethreshold vector array mistake
Status | CLOSED DUPLICATE |
Reported | 2015-02-21 05:13:00 +0100 |
Modified | 2019-08-10 12:31:21 +0200 |
Product: | FieldTrip |
Component: | forward |
Version: | unspecified |
Hardware: | All |
Operating System: | All |
Importance: | P5 normal |
Assigned to: | |
URL: | |
Tags: | |
Depends on: | |
Blocks: | |
See also: |
John Richards - 2015-02-21 05:13:10 +0100
In volumethreshold.m % m(cluster) = find(tmp==k, 1, 'last'); % determine the voxel count for each cluster m(k) = find(tmp==k, 1, 'last'); % determine the voxel count for each cluster The cluster is a large array, whereas m is only the size of the n clusters. So the m(cluster) gives an error (only use int index... or something). I replaced this with m(k), where k is the counter for the n clusters. This gets the voxel count for the k cluster in the m(n-cluster) array. This seems to work ok. This error occurs on Windows and Linux. Also, in the same proc there is a print(....'\m'). This does not seem functional; and linux gives a warning that \m is not recognized. I removed this and it worked fine.
Robert Oostenveld - 2015-02-22 14:02:39 +0100
Hi John, I already fixed this recently, the present code does not use "find" any more. Please update to the latest FT version. See revision details below that pertain to the fix. best Robert ------------------------------------------------------------------------ r10238 | roboos | 2015-02-16 22:26:17 +0100 (Mon, 16 Feb 2015) | 2 lines bugfix - fixed regression error in volumethreshold, see bug 1954#c23. Also improved the segmentation in the test script, it was not good at the bottom of the head. ------------------------------------------------------------------------ r10225 | roboos | 2015-02-12 10:29:27 +0100 (Thu, 12 Feb 2015) | 2 lines bugfix - some small fixes and updates to the test script to get it to work. See http://bugzilla.fcdonders.nl/show_bug.cgi?id=1954 *** This bug has been marked as a duplicate of bug 1954 ***