[daip] hi

[Eric Greisen]

Anish Roshi wrote:
> Hi Eric,
> 
> Is it possible to create a cube of gaussian models using the gaussian 
> amp, central vel and width images produced by xgaus ?
> 
> I am having trouble with 2 component Gaussian fitting on my data. I used 
> doslice = 2 and suppressed one component using the option `e' and 
> putting the cursor outside the window whenever only 1 line component was 
> present in the data. In the display the fit was good (both single 
> component and at pixels where 2 components were present.) But in the 
> output image of one of the components (which is the one I was 
> suppressing whenever I needed only 1 component) that amplitude went 
> crazy -- too high values at several places etc.
> 
> So I am thinking to subtract the gaussian model cube of the line 
> component which gave the good amplitude, width and center in the xgaus 
> output from the original data and then fit the second component alone. 
> Is it possible ?
> 
> Anish


NEVER NEVER send a message with "Hi" as the subject.  Most SPAM filters 
delete them with the user never seeing them.

You seem to be using the new XGAUS so it is possible to re-examine each 
pixel that has bad results when you reach the imaging/editing stage. 
What happens?  Are you using the absolutely latest versions of XGAUS - I 
did have some issues like this but I thought they had been corrected. 
XGAUS shows the numbers it is about to store during the approval stage - 
if they are bad you should be able to deal with them.

I do not know what you mean by cube in this case.  The images of peak 
value, center location, and width have different units and are of 
different things and so do not make a "cube" in any natural sense.

E-mails sent to me personally cannot be addressed when I am away.  Use 
daip at nrao.edu and you may be able to get more rapid response.

Eric Greisen