Hi Troels,
That's a good idea, feel free to implement it. I would suggest coming
up with a different user function name though, as this is 100%
independent of the relaxation dispersion analysis. It might fit
better in the spectrum user function class
(spectrum.deselect_sn_ratio). Or maybe even better, the select and
deselect user function classes (select.sn_ratio, deselect.sn_ratio).
For consistency, it would be better to use all lower case. I would
also suggest creating the spectrum.sn_ratio user function to
pre-calculate the S/N ratio, as this could have other uses. It could
be stored in spin.sn_ratio as a dictionary with the same keys as
spin.intensity and spin.intensity_err.
How would this be implemented though? Would you look at all peak
intensities for a spin divided by the intensity error and deselect if
any is below? Or if all are below?
Regards,
Edward
On 5 December 2014 at 16:25, Troels Emtekær Linnet
<tlinnet@xxxxxxxxxxxxx> wrote:
Hi Edward.
Would it be useful to add additional features on top of the
relax_disp.insignificant function?
A possibility to only select spins, which have a S/N above 10 or more?
This could be run before analysis, and only depends
on spectrum.error_analysis().
The relax_disp.insignificant function is not enough.
If a "weak" peak is loaded, with intensity near the noise, then
calculation
of R2eff can fluctuate quite a bit.
This will not be observable with relax_disp.insignificant.
The SN ratio could even be calculated and stored when doing the
spectrum.error_analysis().
Best
Troels
_______________________________________________
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: Add the userfunction relax_disp.select_SN_ratio for the possibility to deselect spins, with a certain signal/noise ratio