Back to the main page.
Bug 1320 - make test scripts for the pieces of code where spm8 calls ft functions
| Status | CLOSED FIXED |
| Reported | 2012-02-08 16:30:00 +0100 |
| Modified | 2019-08-10 12:04:00 +0200 |
| Product: | FieldTrip |
| Component: | core |
| Version: | unspecified |
| Hardware: | PC |
| Operating System: | Mac OS |
| Importance: | P3 normal |
| Assigned to: | Vladimir Litvak |
| URL: | |
| Tags: | |
| Depends on: | |
| Blocks: | |
| See also: |
Robert Oostenveld - 2012-02-08 16:30:51 +0100
spm8 makes use of low-level and high-level fieldtrip functions. Esp. for the high-level functions it would be good to have a test case to ensure that ft changes don't break the spm8 usage. (as discussed with Guillaume)
Robert Oostenveld - 2012-02-08 16:31:57 +0100
of course it should also be done for spm12
Vladimir Litvak - 2012-07-19 17:55:47 +0200
Hi Robert, I've just checked in a script and a mat file with some typical input for testing some aspects of SPM-FT integration. I might add to it in the future. The script fails in one place for MEG concentric spheres. It has something to do with plotting and I couldn't exactly figure it out. Strangely there is no problem in SPM itself. The only difference is that in SPM everything is in CTF coordinates and in the script in MNI coordinates but in both cases in mm so please take a look in case it's something important. Other than that it should work. We upgraded FT in internal SPM8 and hopefully will keep synced from now on. Best, Vladimir
Vladimir Litvak - 2012-07-27 11:28:41 +0200
Hi Robert, I did what I was supposed to do so tell your bugzilla to stop nagging me now. Vladimir
Vladimir Litvak - 2012-08-02 13:34:39 +0200
Hi Robert, The problem with plotting I mentioned does occur in SPM after all and prevents the multiple spheres model from being useful. Could you run my script and fix it? Thanks, Vladimir
Robert Oostenveld - 2012-11-26 12:15:03 +0100
the test script is test_spm_ft_integration.m, at this moment it fails to run
Robert Oostenveld - 2012-11-30 09:03:36 +0100
after some changes over the last few days it now runs. @Vladimir, could you make a separate script for spm8 and for spm12? The spm8 version would remain fixed from now on and should reflect the use of FT functions in the release version of spm8. The spm12 version would still be flexible and would follow the development in the SPM trunk.
Jörn M. Horschig - 2013-04-23 17:17:19 +0200
btw, that has been failing for longer already: http://fieldtrip.fcdonders.nl/development/dashboard
Jörn M. Horschig - 2013-04-23 17:19:11 +0200
whoopsy, wrong bug, my apologies, this meant to go to bug 2111. but it also belongs here, and Vladimir also got the message, so vladimir, please regard this in conjunction with bug 2111. all others: I just meant to say that the test script Robert was mentioning is failing because of something in ft_prepare_layout going wrong ;)
Robert Oostenveld - 2013-04-23 17:37:46 +0200
(In reply to comment #7) After keeping the test_ft_spm_integration script running for a bit longer I end up with Warning: padding not sufficient for requested frequency resolution, for more information please see the FAQs on www.ru.nl/neuroimaging/fieldtrip > In ft_specest_mtmconvol at 122 In test_spm_ft_integration at 151 Warning: output frequencies are different from input frequencies > In ft_specest_mtmconvol at 128 In test_spm_ft_integration at 151 Error using test_spm_ft_integration (line 154) Assertion failed. 154 assert(length(unique(diff(freqoi)))==1) K>> unique(diff(freqoi)) ans = 0.6663 0.6663 0.6663 0.6663 0.6663 0.6663 0.6663 1.3325 1.3325 1.3325 1.3325 1.3325 1.3325 1.3325 1.3325 At the respective place in the code it is asking for K>> 1:48 ans = Columns 1 through 34 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 Columns 35 through 48 35 36 37 38 39 40 41 42 43 44 45 46 47 48 but getting back K>> freqoi freqoi = Columns 1 through 20 1.3325 1.9988 3.3313 3.9976 5.3301 5.9964 7.3289 7.9952 9.3277 9.9939 11.3265 11.9927 13.3253 13.9915 15.3240 15.9903 17.3228 17.9891 19.3216 19.9879 Columns 21 through 40 21.3204 21.9867 23.3192 23.9855 25.3180 25.9843 27.3168 27.9830 29.3156 29.9818 31.3144 31.9806 33.3131 33.9794 35.3119 35.9782 37.3107 37.9770 39.3095 39.9758 Columns 41 through 48 41.3083 41.9746 43.3071 43.9733 45.3059 45.9721 47.3047 47.9709 This is something I believe Roemer (now CC) recently has been working on.
Robert Oostenveld - 2013-10-30 11:21:04 +0100
the test script exists