Hi Edward, Thank you for your comments. The "relax_data.read" actually included reading the data and errors (so much for copy-pasting). I have updated the relax version to rev. 14791 and the error message now is "UnboundLocalError: local variable 'spin_id' referenced before assignment". I have submitted a bug report with an example attached (#18790).
Vitaly On Wed, Oct 5, 2011 at 04:03, Edward d'Auvergne <[email protected]> wrote: > Hi Vitaly, > > The problem here is not to do with the spin numbering. That is not > needed. I've now updated relax to print out the full spin ID string > instead of just the spin number and name. I've also modified the > error messages to be more informative. Could you update your > repository copy of relax and send the new error messages? Thanks. > One problem I see, though this may not be the cause of the error, is > in the reading of the relaxation data. The script line: > > relax_data.read(ri_id='R1_600', ri_type='R1', frq=600.484*1e6, > file='data.r1.600', mol_name_col=1, res_num_col=2, res_name_col=3, > spin_name_col=5) > > should probably be: > > relax_data.read(ri_id='R1_600', ri_type='R1', frq=600.484*1e6, > file='data.r1.600', mol_name_col=1, res_num_col=2, res_name_col=3, > spin_name_col=5, data_col=6, error_col=7) > > Specifying the data and error columns is very important. Though > according to the code, if the data and error were not read in, then > the checks resulting in the RelaxError should not occur. The problem > with the spin numbers is that the numbers in 'data.r1.600' do not > match the numbers in the first column of the PDB file. For proteins, > it is safe to not specify the spin number as all atoms of the residue > should have a different name. For small organic molecules, RNA, DNA, > sugars, etc, this is not the case. Anyway, the spin numbers in the > structural file and data files must match if they are used at all. > > If you could, submitting a bug report would be appreciated > (https://gna.org/bugs/?func=additem&group=relax). This helps other > relax users if they encounter the same issue. If you could attach a > minimal script with a minimal data set (relaxation data for one or two > residues, randomised if you prefer), then I'll be able to quickly > reproduce the bug and normally fix it within 5 minutes. I would then > also be able to add the script and data to the relax test suite to > make sure that the bug never returns. > > Cheers, > > Edward > > > > > On 5 October 2011 01:48, V.V. <[email protected]> wrote: >> Hello, >> >> It looks like Relax scripts changed a bit since the last time I used >> it. I am trying to setup the initial test calculation with >> "single_model.py", and running into problems with spins initialization >> (revision 14778). >> >> Sample input files: >> ===== >> #mol_name res_num res_name spin_num spin_name data error >> ave 1 Met 1 N None None >> ave 2 Glu 2 N 1.8903 0.06062 >> ... >> ===== >> >> Script extract is below: >> ===== >> name = 'tm1' >> pipe.create(name, 'mf') >> >> sequence.read('data.r1.600', mol_name_col=1, res_num_col=2, >> res_name_col=3, spin_name_col=5) >> structure.read_pdb('ave.pdb', read_mol=1, set_mol_name='ave', >> parser='internal') >> relax_data.read(ri_id='R1_600', ri_type='R1', frq=600.484*1e6, >> file='data.r1.600', mol_name_col=1, res_num_col=2, res_name_col=3, >> spin_name_col=5) >> >> #More data sets are loaded and other parameters initialized >> Grid_search(inc=11) >> ===== >> >> This returns an error: >> "RelaxError: Negative error for spin 'None N', minimization not possible" >> >> Is this because the "spin_number" has not been initialized during the >> "sequence.read" and "relax_data.read"? If I try to use >> "spin_num_col=4" in "sequence.read", then "structure.read" returns an >> error: "RelaxWarning: Cannot find the atom in the structure (atom ID >> ':2@2&@N')" >> >> Please let me know if you need more details. >> >> Thank you, >> Vitaly >> >> _______________________________________________ >> relax (http://nmr-relax.com) >> >> This is the relax-users mailing list >> [email protected] >> >> 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 >> > _______________________________________________ relax (http://nmr-relax.com) This is the relax-users mailing list [email protected] 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
