iraf.net

[b:c7440c41be]From: [/b:c7440c41be]"safi khalid" <astro20024@lycos.com ([email]astro20024@lycos.com[/email])>
[b:c7440c41be]Date: [/b:c7440c41be]December 19, 2009 11:14:42 PM MST
[b:c7440c41be]To: [/b:c7440c41be]<valdes@noao.edu ([email]valdes@noao.edu[/email])>
[b:c7440c41be]Subject: [/b:c7440c41be][b:c7440c41be] IRAF Help[/b:c7440c41be]
Dear professor Valdes ,

Iam sorry for disturbing you. But I want to learn things but find no help
Iam a user of IRAF (not expert!). a student starting master.
I would like you help me with finding a way to do a statistical thing
with IRAF:
well, I have a set of curves (x and y points connected together).
I converted the dats such to have a fits format version of the curves
and then I can work on them like spectra..... what
I want to do is smooth all curves in the same way and then find
an average (or median...I dont know what is the best and how tod do that?)
any way. I want find a curve that can represent the set of curves. like if Iam
looking for a template that may represent my curves.
I found a command called 'scombine'; I dont know if that is the right
command to work with, or better others. note that some coverage in x-axis
lack in some curves.
then I want find the statistical data of that work (parameters that
can explain the goodness of the template to represent the set of vectors).
In conclusion my questions are:
- how to smooth curves in a same way? and what are parameters of the smooth method?
- How to find an average/median of them (like a representative of the set of curves)?
- Any statistical (meaningful parameters) that can be discussed from that?

I would be so so gratefull if you can help me with this.
many thanks for your efforts and support

sorry for disturb again
safi [b:c7440c41be]From: [/b:c7440c41be]Frank Valdes <valdes@noao.edu ([email]valdes@noao.edu[/email])>
[b:c7440c41be]Date: [/b:c7440c41be]December 23, 2009 11:18:39 AM MST
[b:c7440c41be]To: [/b:c7440c41be]"safi khalid" <astro20024@lycos.com ([email]astro20024@lycos.com[/email])>
[b:c7440c41be]Subject: [/b:c7440c41be][b:c7440c41be]Re:  IRAF Help[/b:c7440c41be]
Hello Safi,
If you have your curves as 1D images you can use the tasks in the imfilter package:
         boxcar - Boxcar smooth a list of 1 or 2-D images
       convolve - Convolve a list of 1 or 2-D images with a rectangular filter
        fmedian - Quantize and box median filter a list of 1D or 2D images
          fmode - Quantize and box modal filter a list of 1D or 2D images
       frmedian - Quantize and ring median filter a list of 1D or 2D images
         frmode - Quantize and ring modal filter a list of 1D or 2D images
          gauss - Convolve a list of 1 or 2-D images with an elliptical Gaussian
       gradient - Convolve a list of 1 or 2-D images with a gradient operator
        laplace - Laplacian filter a list of 1 or 2-D images
         median - Median box filter a list of 1D or 2D images
           mode - Modal box filter a list of 1D or 2D images
        rmedian - Ring median filter a list of 1D or 2D images
          rmode - Ring modal filter a list of 1D or 2D images
         runmed - Running median a list of images at each pixel position
In particular, you might just use BOXCAR which is a running average.
To make a template the task IMCOMBINE would be appropriate.  This also works
on all types of images.  The task SCOMBINE is specific to spectra and can be
used but this requires having a coordinate system.  If things are just basically
in spatial units of pixels then IMCOMBINE and SCOMBINE do the same
thing.
Yours,
Frank Valdes
[b:c7440c41be]From: [/b:c7440c41be]"safi khalid" <astro20024@lycos.com ([email]astro20024@lycos.com[/email])>
[b:c7440c41be]Date: [/b:c7440c41be]December 28, 2009 7:22:33 AM MST
[b:c7440c41be]To: [/b:c7440c41be]"Frank Valdes" <valdes@noao.edu ([email]valdes@noao.edu[/email])>
[b:c7440c41be]Subject: [/b:c7440c41be][b:c7440c41be]Re:  IRAF Help[/b:c7440c41be]

Dear Professor Valdes, Thank you very very much for your kindness in replying and helping
me. I appreciate that and you a lot.
I am trying the commands. I have just one more question, if you allow me:
how can I know the parameters of the goodness of the smooth and the combine.
I think it should be a way to see the results of the smoothing process
and also the combine; I mean like the parameters used to do the tasks.
like for example in case I make a fit for example a code should give me the
chi2 in order to see how good is that fit.
I dont know how to visualize and figure out the parametrs used by the
above tasks to give an out put result, since it gives me just outputs.

many thanks and sorry sorry once again
HAPPY NEW YEAR OF PEACE AND HEALTH AND SUCCESS
SAFI
[b:c7440c41be]From: [/b:c7440c41be]Frank Valdes <valdes@noao.edu ([email]valdes@noao.edu[/email])>
[b:c7440c41be]Date: [/b:c7440c41be]January 13, 2010 4:46:47 PM MST
[b:c7440c41be]To: [/b:c7440c41be]safi khalid <astro20024@lycos.com ([email]astro20024@lycos.com[/email])>
[b:c7440c41be]Subject: [/b:c7440c41be][b:c7440c41be]Re:  IRAF Help[/b:c7440c41be]
Hello Safi,You will have to figure this out on your own, possibly by scripting some image arithmetic steps.  The tasks do not provide goodness of fit unfortunately.
Happy New Year
Frank