Re: Errors of the 'final' Mode of Model Free AnalysisRe: Errors of the 'final' Mode of Model Free AnalysisHi Jing, I have now created a bug report for the problem (https://gna.org/bugs/?21079). This is a real bug in relax, but I cannot fix it due to not being able to replicate the problem and not having enough information to determine where the problem could be. The "local tm" global diffusion model is a perfectly valid model. It is useful for IDPs, multi-domain systems, flexible systems, or anything that does not diffuse as a spherical, spheroidal, or ellipsoidal rigid geometric object. The selection of this model indicates that these basic rigid geometric descriptions of diffusion are not suitable for your system. The protocol should then continue to completion and not show this RelaxError! Alternatively the selection of the "local tm" global diffusion model can indicate that your data is not consistent. Have you tried Sebastien Morin's consistency testing analysis in relax? See http://www.nmr-relax.com/refs.html#Morin09 for details. Have you collected your data with proper temperature calibration and control? I.e. you need to calibrate each experiment on each spectrometer with MeOH or ethylene glycol. And you need a temperature compensation block at the start of your pulse sequence and/or single scan interleaving. See http://www.nmr-relax.com/manual/Temperature_control_calibration.html, http://www.nmr-relax.com/manual/relax_data_temp_calibration.html, and http://www.nmr-relax.com/manual/relax_data_temp_control.html. I hope this information helps. I will try again to replicate the bug you see, but I have already failed multiple times. relax results files attached to that bug report would make things much easier for me. Cheers, Edward On 28 August 2013 17:49, 赵竞 <jzhao16@xxxxxxxx> wrote:
Hi,Edward: Thanks for the information you provided. I think the reason why Nicolas and I have this error message is because our local_tm model has the lowest chi 2. So the program picked up the local_tm model. However, local_tm doesn't belong to the model from (m1 ~ m9), so no specific has been selected. It is kind of weird because the chi 2 would be larger after the program fits the data with more specific diffusion model. Do you know if there is a way that I can fix this problem to run the "final" model and have my model selected. Thanks Jing 2013/8/20 Edward d'Auvergne <edward@xxxxxxxxxxxxx>Hi Jing, Welcome to the relax mailing lists! The error you see has been reported before by Nicolas Doucet, see the email thread at: http://thread.gmane.org/gmane.science.nmr.relax.user/1299 Unfortunately in that case the problem disappeared without the reason being found. But I would recommend reading my message at the top of that thread as I can only repeat what I said there - check the log messages earlier up. Submitting a bug report would be very useful in solving this problem, as log and other files can be attached there and I can track and handle the issue (https://gna.org/bugs/?func=additem&group=relax). You could also upgrade to relax 3.0.0 as that fixes a number of bugs, though I don't think that this issue has been resolved (https://gna.org/forum/forum.php?forum_id=2407, http://article.gmane.org/gmane.science.nmr.relax.announce/43). Regards, Edward On 20 August 2013 07:42, Jing Zhao <jzhao16@xxxxxxxx> wrote:Hi,Edward: I have already used the program relax 2.2.5 to run through all diffusion model: Sphere, Prolate, Oblate and Ellipsoid in a uni-processor and obtained the converged fitting results that passed the convergence test. However, I stuck on the "final" mode because of some errors as shown below: relax> value.set(val=-0.00017199999999999998, param='csa', spin_id='@N') relax> pipe.create(pipe_name='local_tm - mf', pipe_type='mf', bundle='mf') relax> results.read(file='results', dir='C:\\relax\\ Model Free Analysis\\local_tm\\aic') Opening the file 'C:\\relax\\Model Free Analysis \\local_tm\\aic\\results.bz2' for reading. relax> pipe.create(pipe_name='sphere - mf', pipe_type='mf', bundle='mf') relax> results.read(file='results', dir='C:\\relax\\Model Free Analysis \\sphere\\round_13\\opt') Opening the file 'C:\\relax\\Model Free Analysis \\sphere\\round_13\\opt\\results.bz2' for reading. relax> pipe.create(pipe_name='prolate – mf', pipe_type='mf', bundle='mf') relax> results.read(file='results', dir='C:\\relax\\Model Free Analysis \\prolate\\round_123\\opt') Opening the file 'C:\\relax\\Model FreeAnalysis \\prolate\\round_123\\opt\\results.bz2' for reading. relax> pipe.create(pipe_name='oblate - mf', pipe_type='mf', bundle='mf') relax> results.read(file='results', dir='C:\\relax\\ Model Free Analysis\\oblate\\round_14\\opt') Opening the file 'C:\\relax\\Model Free Analysis \\oblate\\round_14\\opt\\results.bz2' for reading. relax> pipe.create(pipe_name='ellipsoid - mf', pipe_type='mf', bundle='mf') relax> results.read(file='results', dir='C:\\relax\\Model Free Analysis \\ellipsoid\\round_912\\opt') Opening the file 'C:\\relax\\Model Free Analysis \\ellipsoid\\round_912\\opt\\results.bz2' for reading. relax> model_selection(method='AIC', modsel_pipe='final - mf', bundle='mf', pipes=['local_tm - mf ', 'sphere - mf', 'prolate - mf', 'oblate - mf', 'ellipsoid - mf ']) AIC model selection. Global model - all diffusion tensor parameters and spin specific model-free parameters. # Data pipe Num_params_(k) Num_data_sets_(n) Chi2 Criterion local_tm - mf (Mon Aug 19 22:40:08 2013) 323 696 5983.71348 6629.71348 sphere - mf (Mon Aug 19 22:40:08 2013) 237 696 6374.27910 6848.27910 prolate - mf (Mon Aug 19 22:40:08 2013) 232 696 6209.77670 6673.77670 oblate - mf (Mon Aug 19 22:40:08 2013) 235 696 6274.97999 6744.97999 ellipsoid - mf (Mon Aug 19 22:40:08 2013) 238 696 6184.98839 6660.98839 The model from the data pipe 'local_tm - mf' has been selected. relax> monte_carlo.setup(number=500) relax> monte_carlo.create_data(method='back_calc') RelaxError: The specific model has not been selected or set up. Do you have any idea what is wrong in the here? Thanks Jing _______________________________________________ relax (http://www.nmr-relax.com) This is the relax-users mailing list relax-users@xxxxxxx To unsubscribe from this list, get a password reminder, or change your subscription options, visit the list information page at https://mail.gna.org/listinfo/relax-users-- Zhao, Jing | 赵竞 Partners III, Room 363 851 Main Campus Drive Department of Chemistry, North Carolina State University. Tel : 919-720-0926