autoidenify does not auto find features, but can with “
holt wrote on Jun 25, 2009
Hi,
I am trying to have autoidentify run automatically for our tests. There are four features in my arc and I provided a coordlist with wavelengths of said four features. Initially autoidentify does not find a solution; however, in interactive mode, all I need to do is press "l" to match the features with the linelist and it works beautifully. The features are marked and the solution looks good. What does "l" do that the intial algorithm does not? Are there parameters that I can set to mimic "l" in non-interactive mode?
Thank you very much for your time.
-JH
I am trying to have autoidentify run automatically for our tests. There are four features in my arc and I provided a coordlist with wavelengths of said four features. Initially autoidentify does not find a solution; however, in interactive mode, all I need to do is press "l" to match the features with the linelist and it works beautifully. The features are marked and the solution looks good. What does "l" do that the intial algorithm does not? Are there parameters that I can set to mimic "l" in non-interactive mode?
Thank you very much for your time.
-JH
Francisco Valdes wrote on Jun 25, 2009
AUTOIDENTIFY requires fairly linear spectra (wavelenth=A+B*pixel) and a reasonable number of features. What it does in the automatic mode is search for patterns of lines (triplets, quartets, etc). So if your arc only has four lines this would almost certainly not work.
The 'l' key in the IDENTIFY/AUTOIDENTIFY must have a starting wavelength solution from which it can automatically add lines from a line list. If you have a totally uncalibrated spectrum it would not work.
There are ways to automate line identifications but it depends on what information is know a priori or how general a solution is needed. If you have high resolution data with only a few lines then you must be able to supply some intial information about the dispersion and central wavelength.
If you will provide more information I might be able to give you some ideas.
Yours,
Frank Valdes
The 'l' key in the IDENTIFY/AUTOIDENTIFY must have a starting wavelength solution from which it can automatically add lines from a line list. If you have a totally uncalibrated spectrum it would not work.
There are ways to automate line identifications but it depends on what information is know a priori or how general a solution is needed. If you have high resolution data with only a few lines then you must be able to supply some intial information about the dispersion and central wavelength.
If you will provide more information I might be able to give you some ideas.
Yours,
Frank Valdes
Last post on Jun 25, 2009