Hi,
This should be possible. It is only a question of difficulty. The
infrastructure is not in place to handle this as only the
back-calculated values corresponding to the measured values are
calculated. This is more scientifically valuable as the quality of
fit should be judged by the residuals, specifically if there is a bias
in the residuals.
However having a graph with more points can be added. I am assuming
that the aim is the individual graphs produced by the
relax_disp.plot_disp_curves user function. This would be the easiest
place to add such a feature. If you are interested in the grace.write
user function, we will be in much, much more trouble as such
interpolation cannot be made general, especially if the x-axis are the
residue numbers.
The way this could be done, assuming we are working with the
relax_disp.plot_disp_curves user function, is to use the target
function code already in relax. I would suggest adding this as
another set to the graphs, leaving all the current sets as they are.
Then it can be turned on and off, as desired. For inspiration, have a
look at the _back_calc_r2eff() method in
specific_analyses.relax_disp.api. The key would be to mimic this
method but increase the dimensionality of the data structures for the
interpolation. The values, errors and missing data structures can be
created with the numpy.zeros and numpy.ones functions. You will also
need to increase the dimensionality of the cpmg_frqs and spin_lock_nu1
structures sent into the Dispersion target function class for the
interpolation - both will have to be handled! Hence both would need
to be tested in the test suite. Have a look around the code and see
what you think.
Regards,
Edward
On 9 September 2013 21:47, Troels Emtekær Linnet <tlinnet@xxxxxxxxx> wrote:
Hi Edward.
I would like to produce some graphs with more points than the standard
graphs.
Particularly, I am looking for something similar to make an numpy
arange from min to max of cpmg frequencies, and interpolate with 50
points?
Calculate the R2eff values from the fitted parameters of the model equation
at populate a list of y_values.
That would produce more interesting graphs to look at, than graphs
with only 10-20 points.
Is there a way to call the model function?
Troels Emtekær Linnet
_______________________________________________
relax (http://www.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
Re: Is there a way to send an array of CPMG values into target function?