Subject: Re: [nmr_sparky] Peak Selection
From: Mandar T. Naik
Date: Jul 20, 2006

Previous: 192 Next: 194


Wow! That was a wonderful reply. Thanks.
I generally do it similar to your suggestion 1 and yesterday did spent lot of
time trying a work around relevant to suggestion 3.
But you gave a good tip of decreasing the negative contour level while using
kr, I think that minimized my problem to a considerable extent. I was under
the impression that once you keep those at 0, the level height is immaterial,
it seems I was wrong.
Best regards
-mandar


On Thursday 20 July 2006 17:55, e.ab@... wrote:
Quoting Mandar T. Naik mandarn@... :
Hi Thomas, Eiso and Everybody out there,

Does anyone know a trick or can suggest code for selecting negative data
height peaks? Or sort the peak list, lt with respect to data heights (I
know it has a option for peak volumes).

Best regards
-mandar

there are several clumsy ways.
1. write out a peaklist with data heights, remove the negative peaks
in an editor, and read in the peaklist again (and remove your old peaks)
you will lose your peak label positions.

2. do integration with a fixed, very small fixed linewidth (so that
effectively
only the region around one point is integrated.) maybe a small box size
is also possible. then use lt and sort on volume. swith off all
optimisation options for the integration to make it go fast.


3. hack the LT peak-dialog code so you can sort on data-height. Ive have
done this but I cant access it right now (vacation). this dialog however
will not allow you to select more than one peak at the time, so its still
tedious.

4. open a copy of your spectrum and use the old spectrum to do restricted
peakpicking on the new spectrum with the negative contour very low.

5 a variation of 1.) which keeps the label positions: open a copy of
your spectrum and read in a peaklist with only the negative peaks. now
use
the restricted peak picking dialog to _select_ only the negative peaks
in the original spectrum and delete them.

Eiso