Back to the main page.
Bug 2801 - calling ft_stratify gives error when too many Curve Fitting Toolbox licenses are used
| Status | CLOSED WONTFIX |
| Reported | 2015-01-06 16:15:00 +0100 |
| Modified | 2016-11-29 09:03:05 +0100 |
| Product: | FieldTrip |
| Component: | core |
| Version: | unspecified |
| Hardware: | PC |
| Operating System: | Windows |
| Importance: | P5 enhancement |
| Assigned to: | Jan-Mathijs Schoffelen |
| URL: | |
| Tags: | |
| Depends on: | |
| Blocks: | |
| See also: |
Stan van Pelt - 2015-01-06 16:15:59 +0100
When calling ft_stratify multiple times simultaneously (e.g. when using torque), it returns an error when too many Curve Fitting Toolboxes are used. This seems to be caused by calling of randperm.m.
Robert Oostenveld - 2015-01-06 17:57:05 +0100
This is not really a bug. Just a limit of the licenses you have available. Note that randperm does not seem the culprit. >> which randperm built-in (/Applications/MATLAB_R2014b.app/toolbox/matlab/randfun/randperm) If you can find the function that is from the curve fitting toolbox, we might try to replace that with our own and not rely on an external toolbox.
Jan-Mathijs Schoffelen - 2015-01-07 08:54:09 +0100
Hoi Stan, I am fairly sure that randperm is *not* a function in the curve fitting toolbox, so I doubt that your diagnosis of the problem you encountered is correct :-). You are not by any chance doing any fancy gamma frequency peak detection in your matlab-script, fitting gaussians to spectra? Note that yesterday there where some issues with the stability of the torque cluster, which by now should have been resolved. I am not sure, but this may have been a cause, too. Could you try again and get back to us if your problem persists, otherwise close the issue?
Stan van Pelt - 2015-01-07 10:36:58 +0100
JM, you might be right that it had to do with torque. I got the error before the torque-rollback yesterday. I could not copy the exact error unforunately, because my session was terminated due to the version-rollback. Today, I cannot replicate the error, so as for now, I will close the bug.
Stan van Pelt - 2015-01-07 11:47:28 +0100
License checkout failed. License Manager Error -4 I've reopened the bug, because I am getting the same error again. To me, it seems to be caused by randperm? Maximum number of users for Curve_Fitting_Toolbox reached. Try again later. To see a list of current users use the lmstat utility or contact your License Administrator. Troubleshoot this issue by visiting: http://www.mathworks.com/support/lme/R2012b/4 Diagnostic Information: Feature: Curve_Fitting_Toolbox License path: 1700@fcdc006s:/home/mrphys/stavpel/.matlab/R2012b_licenses:/opt/matlab/R2012b/licenses/license.dat:/opt/matlab/R2012b/licenses/network.lic Licensing error: -4,132. Error in ft_stratify (line 259) tmpsel = randperm(length(indone)); Error in bow_grangerTPJSTS (line 91) [strat] = ft_stratify(cfg, SNR_TPJ', SNR_STS'); Error in fexec (line 156) feval(fname, argin{:}); Error in qsubexec (line 89) [argout, optout] = fexec(argin, optin);
Robert Oostenveld - 2015-01-07 13:41:25 +0100
(In reply to Stan van Pelt from comment #4) it still ain't a FT problem. I don't understand why it looks at fcdc006s for a license, and don't understand why it complains about the curve fitting toolbox. It might be that the license server is incorrectly configured. I suggest you report it to the TG.