Subject: Re: [nmr_sparky] Integration by boxing
From: Tom Goddard
Date: Apr 13, 2010

Previous: 669 Next: 671

Hi Gary,

The square outline for selected peak markers is the same size as the X shaped markers. If I make the X peak markers smaller the square box around them when they are selected gets smaller. So I dont understand your distinction between the peak marker size and the square selection box size.

Please post replies to the NMR Sparky Yahoo group so everyone can benefit from the answers.

Tom

Hi Tom,

I did try to make the peak marker smaller, but it didnt seem to be the
peak marker itself that mattered.  I thought it might be a contour line
thing at first, but that wasnt it.

It *might* be the box that is drawn around the peak marker when it is
selected.  I made sure that no peaks were selected when I did the
boxing, but the act of boxing always selected the neighboring peak.  I
think that if the mouse enters the area in which the box would occupy
(if it was there), it automatically selects it, too.

(Did that make sense?)

Thanks,
Gary


On Tue, 2010-04-13 at 12:30 -0700, Tom Goddard wrote:

Hi Gary,

  When you drag a box to integrate by boxing, all the peaks in the box 
do get their volume set to the same value.  In testing with Sparky 3.115 
I do not see it integrating peaks outside the dragged box like you 
report.  Ah, actually I see that if the dragged box contains any part of 
the peak marker, even just a corner, then it gets included.  So the 
center of the peak doesnt actually have to lie inside the dragged box.  
This is the same behavior with peak selection.  Doesnt seem so good.  
You might want to make your peak markers smaller (ornament tool, ot) if 
this is causing problems in overlapped regions.

  Sparky hasnt drawn ellipses in integrate by ellipse mode since about 
1995 when the code was rewritten to run on machines other than Sparc 
stations.

    Tom

Hi Tom,

Im trying to integrate a 2D plot in Sparky using the boxing method, but
am running into some troubles.  It seems that whenever two peaks are
adjacent to each other, the act of integrating a neighbor causes its
near neighbors to be merged or overlapped, and their integrals are set
to the same value.  If I lock the peak positions first, they dont
physically move, but they are given equal integrals.

It used to be that boxing would just set the peak position to the center
of the box (or ellipse) but it would never move a peak that was outside
of the box.  It was very useful that way.

How can I do this?

Also, in the olden-days, the ellipsoid integration would show the
ellipse that you are drawing.  It doesnt seem to do that anymore.  

I am running Sparky 3.113 on Ubuntu, if that matters.

Any help would be appreciated.

Gary