Welcome to iraf.net Monday, May 13 2024 @ 08:18 AM GMT
rohit |
05/04/2009 06:23PM (Read 1779 times)
|
|
|
Status: offline
Registered: 06/01/2006
Posts: 86
|
I have used ecidentify and ecreidentify for high-res IR spectra. I would like to compute the wavelength dispersion error that I get from wavelength solution. Could you please let me know how to find that for each order?
|
|
|
|
valdes |
05/04/2009 06:23PM
|
|
|
Status: offline
Registered: 11/11/2005
Posts: 728
|
Hello and sorry for not responding sooner.First I'm not sure what statistic you are thinking of. One would be the rms of the difference between the computed wavelength from the dispersion function and the true wavelengths (the "user" wavelengths). The only place where this appears for you is when you interactively do the fit in ecidentify, the 'f' key, the title section shows an rms value. Note that since the solution performed by ecidentify is a 2D function over all the lines this is a measure of the dispersion function over all orders.If you want this for individual orders there are two approaches. One is to use identify and simply treat each order as an independent spectrum. To get the rms it would be basically the same thing of looking at the output of the fit. The other is to look at the database file directly and compute or write a script to compute things from the file. The file is normally database/ec<imagename>.ec. You would get the lines for the particular order (either column 1 or column 2) and then compute the error based on column 4 (the computed value) and column 5 (the "true" wavelength).If you are thinking of some other statistic you would, again, have to resort to the information found in the database file which also has the measured pixel position in column 3.All this discussion is mostly about the arcs. You can use ecidentify on an object spectrum if you can mark lines (either emission or absorption) with known wavelengths.I hope this helps.Yours,
Frank Valdes
|
|
|
|
| |
|
Content generated in: 0.06 seconds |
|