Re: Hybrid selection problem ... -- November 10, 2006 - 09:41

That script looks almost identical to the script I used for
hybridising multiple diffusion tensors on a single molecule and then
using AIC model selection to select between the hybrid and the full
tensor.  Although this bug has been triggered (I'm guessing here so I
could be wrong) by improper use of the model_selection() user function
in combination with the hybrid model, would you like to create a bug
report for it?

For the model selection of the global diffusion models (the sphere,
oblate spheroid, prolate spheroid, ellipsoid, the local tm models, and
the hybrid models) eash model must consist of exactly the same number
of relaxation data points (n) and exactly the same number of spins
(l).  If one spin is selected in the sphere and deselected in the
ellipsoid - they cannot be compared as l_sphere != l_ellipsoid and
model selection cannot be used to select between them.  In the case of
your script, the number of spins used in the hybrid is the total
number n.  But it is being compared to two models which consist of
incomplete subsets (x < n and n-x < n) of the total number of spins n.
Hence model selection cannot be applied.  If the hybrid consists of
all spins n, it must be compared to other model consisting of all
spins n.  relax should therefore be modified print a RelaxError
message stating that the models cannot be compared, either due to the
number of spins being different or the number of relaxation data sets
being different.

Cheers,

Edward


On 11/10/06, Alexandar Hansen <viochemist@xxxxxxxxx> wrote:
> Hi,
>
> I'm trying to run a hybrid model selection on my data and have been getting
> a Traceback error.  This is my script and the output.
>
> script = 'my_hybrid.py'
> ----------------------------------------------------------------------------------------------------
>
> run.create('Dom1','mf')
> run.create('Dom2','mf')
> run.create('hybrid_select','mf')
> results.read('Dom1','results.bz2','15N_final_domain1/final')
> results.read('Dom2','results.bz2','15N_final_domain2/final')
>
> run.hybridise('hybrid',['Dom1','Dom2'])
> model_selection('AIC','hybrid_select',['Dom1','Dom2','hybrid'])
> ----------------------------------------------------------------------------------------------------
>
> relax> run.create(run='Dom1', run_type='mf')
> relax> run.create(run='Dom2', run_type='mf')
> relax> run.create(run='hybrid_select', run_type='mf')
> relax> results.read(run='Dom1', file='results.bz2',
> dir='15N_final_domain1/final', format='columnar')
> Opening the file '15N_final_domain1/final/results.bz2' for
> reading.
> Loading all structures from the PDB file.
> Structure('etar_au22_phage_df_pt.pdb', model=1):
>   Nucleotide chain of length 68
>
> relax> results.read(run='Dom2', file='results.bz2',
> dir='15N_final_domain2/final', format='columnar')
> Opening the file '15N_final_domain2/final/results.bz2' for
> reading.
> Using the structures from the run 'Dom1'.
> Structure('etar_au22_phage_df_pt.pdb', model=1):
>   Nucleotide chain of length 68
>
> relax> run.hybridise(hybrid='hybrid', runs=['Dom1', 'Dom2'])
> relax> model_selection(method='AIC', modsel_run='hybrid_select',
> runs=['Dom1', 'Dom2', 'hybrid'])
> AIC model selection.
>
> Instance 0.
>
> Run                  Num_params_(k)       Num_data_sets_(n)    Chi2
>        Criterion
> Dom1                 9                    12                   3.70303
>        21.70303
> Dom2                 10                   12                   1.89550
>        21.89550
> Traceback (most recent call last):
>   File "relax", line 454, in ?
>     Relax()
>    File "relax", line 166, in __init__
>     self.interpreter.run()
>   File "/local/home/viochem/relax/prompt/interpreter.py",
> line 213, in run
>     run_script(intro=self.relax.intro_string, local= self.local,
> script_file=self.relax.script_file, quit=1)
>   File "/local/home/viochem/relax/prompt/interpreter.py",
> line 388, in run_script
>     console.interact(intro, local, script_file, quit)
>   File "/local/home/viochem/relax/prompt/interpreter.py",
> line 340, in interact_script
>     execfile(script_file, local)
>   File "my_hybrid.py", line 9, in ?
>
> model_selection('AIC','hybrid_select',['Dom1','Dom2','hybrid'])
>   File
> "/local/home/viochem/relax/prompt/model_selection.py", line
> 121, in model_selection
>
> self.relax.generic.model_selection.select(method=method,
> modsel_run=modsel_run, runs=runs)
>   File
> "/local/home/viochem/relax/generic_fns/model_selection.py",
> line 217, in select
>     k, n, chi2 = self.model_statistics [run](run=run, instance=i,
> global_stats=global_stats)
>   File "/local/home/viochem/relax/specific_fns/hybrid.py",
> line 117, in model_statistics
>     for i in xrange(num):
> NameError: global name 'num' is not defined
>
>
> Is this a bug, or am I doing something totally wrong?
>
>
>
> Alex
>
>
>
> _______________________________________________
> relax (http://nmr-relax.com)
>
> This is the relax-devel mailing list
> relax-devel@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-devel