Re: Software to perform inhomogeneous ensemble photometry

[Michael Koppelman <lolife@bitstream.net>]

Hello sir.

I'm curious about one thing. I'm sure I could find it if I dug through 
your materials so I apologize for that.

I am very interested in uncertainties. Does your software compute an 
uncertainty for each data point? As I recall, when I've looked at your 
photometry, you do not report uncertainties per data point.

The software I use, Mira Pro, gives a 1/SNR type uncertainty for each 
point. It also writes the zeropoint error into the FITS header for each 
image. I can also (of course) calculate the sigma of the comp star 
magnitudes. These three different uncertainty values allow me to go 
nuts computing total uncertainties. I don't know what algorithm Mira 
uses for the ensemble photometry, but it seems to work pretty well. 
When you are done the standard deviation of any given comp star is the 
same as for all the other comp stars. It is basically doing some sort 
of least squares.

Thanks!
Cheers,
Michael Koppelman
http://www.lolife.com/astronomy/



On Nov 8, 2004, at 5:44 PM, Stupendous Man wrote:

>
>   For years, I've been using the inhomogeneous ensemble photometry
> technique described in Honeycutt, PASP 104, 435 (1992)
>
>      http://adsabs.harvard.edu/cgi-bin/nph-bib_query?
>           bibcode=1992PASP..104..435H&db_key=AST&high=416a79eb0026781
>
> to analyze measurements of a set of many stars on many images.
> I have _finally_ managed to put my ancient code into a form that
> might possibly be used by other people.   This is just a first
> attempt, though, and I know of one particular issue which needs
> to be fixed (see below).
>
>   Nonetheless, you can find the software at
>
>        http://spiff.rit.edu/ensemble/
>
> It comes in a standard "configure"-type tarball.  Download the
> *.tar.gz file, un-zip, un-tar, run "./configure" and then "make" ...
> it should be the regular drill.   The package includes a Perl
> script called "selftest.pl" which creates a small dataset and
> tries to run the software on it.  You can run
>
>         make check
>
> to run the selftest script; if all goes well, you'll see a message
> like "All test passed."
>
>   This is the same software I've used to analyze loads of images
> of several cataclysmic variables taken at the RIT Observatory, so
> I'm pretty sure that it _can_ produce good results in some situations.
> You can find examples of its use by looking at any of the
> nights listed at
>
>          http://stupendous.rit.edu/richmond/ritobs/ritobs.html
>
> with observations of V603 Aql, or ASAS 002511, or WZ Sge.
>
>   You can read about the one item I'd like to replace soon by
> looking at the NEWS file.
>
>   What does this have to do with TASS?  Well, I've used this software
> to analyze a batch of Mark IV measurements for stars in some small
> (say, 0.5-by-0.5-degree) area of the sky.  See Tech Note 100:
>
>      http://stupendous.rit.edu/tass/technotes/tn0100.html
>
> I did this "by hand", downloading a set of measurements from the
> Mark IV database and then massaging them to put them in a format
> that my code could read properly.  It took several hours and was
> a real pain.
>
>   I re-wrote a few small sections of my ensemble analysis code
> to make it much easier to run it on data taken from a database.
> I _think_ that it may be possible now to create a script which,
> given some (RA, Dec) position,
>
>           - grabs all data for all stars in the region from the
>                      Mark IV database
>           - breaks up the result into a set of pieces, one
>                      per Mark IV image (just sort by JD and
>                      grab all measurements with the same JD)
>           - runs the ensemble code on the resulting data
>                      to generate improved _relative_
>                      photometry of stars in the region
>
>   I will try doing this myself, but given my current commitments,
> probably won't finish it this calendar year.  Sigh.
>
>   I guess it's possible that some people reading this message
> might want to use the code on non-TASS data, too ...
>
>                                              Michael Richmond
>
>