PSFMeasure returning same radius

Elden Paul Iverson wrote on May 22, 2008

Question for Frank:

PSFMeasure is returning the same radius for differing desired flux percentages. Whether I say I want to know the radius that contains 95% or 99% of the profile the same radius value is returned. Why is this and how can I fix it?

Thanks.

-Paul Iverson

---

Francisco Valdes wrote on May 22, 2008

Hi Paul,

I don't know immediately. Post the parameters and I'll try it on some test data to see if I get the same behavior.

Regards,
Frank Valdes

---

craigswenson wrote on May 22, 2008

Hey Frank,

Paul's not in yet, so I'll follow up on this one. Here are the parameters that we're using in psfmeasure:

(size   =               Radius) Size to display
(beta   =                INDEF) Moffat beta parameter
(scale  =                   1.) Pixel scale
(radius =                   5.) Measurement radius (pixels)
(sbuffer=                   5.) Sky buffer (pixels)
(swidth =                   5.) Sky width (pixels)
(saturat=               65535.) Saturation level
(ignore_=                  yes) Ignore objects with saturated pixels?
(iterati=                   10) Number of radius adjustment iterations
(xcenter=                INDEF) X field center (pixels)
(ycenter=                INDEF) X field center (pixels)
(logfile=              logfile) Logfile

We then vary the "level" parameter to include different percentages of the object in question. The output radius changes as we increase "level" from 90% - 95%, but then as we continue past 95% psfmeasure keeps giving us the same value (the one it gave us for 95%). Originally we had the "iterations" parameter set to 5 (default value) and thought that maybe we weren't allowing it to go through enough iterations to get to the correct answer, so we bumped the iterations up to 10, but still got the same answers. Any ideas?

Thanks,

Craig

---

Francisco Valdes wrote on May 22, 2008

Thanks Craig,

This gave me enough to go on to check the code. The code limits the level to be tween 5% and 95%. I felt that if you pushed to far out then noise and the vagaries of PSF wings would make the measures not very meaning full. If you really want to push that far into where the wings merge into the sky you will have to use some other method such as aperture photometry.

Sorry for the confusion.

Frank

---

craigswenson wrote on May 22, 2008

Thanks Frank.

It's a relief knowing that there is a simple explanation, we'll just stick to nothing larger than 95%.

-Craig

---

Last post on May 22, 2008