The ISAP Line fitting module

First zoom (left-mouse button selection) on required lines, the new window should contain some baseline, then click on one of the Fit buttons. They are three possible patterns :
• Gauss Fit : line is supposed to be fitted by a gauss function (like IDL routine Gaussfit) i.e. on selection
  flux = exp(-((wavelength - line_center)/line_width)^2/2) * line_height + polynomial baseline
• Sinus Fit : line is supposed to be fitted by a sinus function
  
  flux = (sinus(pi * (wavelength - line_center)/line_width + pi/2) + 1) * line_height + polynomial baseline
• Element Fit : line is supposed to be fitted by LWS spectral element.
  The fit is slightly different as the pattern is given on a point by point base by rescaling LWS spectral element curve with line height, line width and line center position, then interpoling on required wavelengths.
  As the resulting function have no obvious partial derivatives, the fit is done estimating them with differences.

All three patterns are fitting 6 parameters :
• 3 parameters for line : line position (wavelength), line height, line width
• 3 parameters for a polynomial baseline of degree 2

When fit is done, baseline and line fit are overplotted on the spectrum (orange and yellow). Then a window is displaid with :
• lines parameters (height, width, center and flux)
• baseline approximation
• list of existing element or molecules who emits close enough to this line (given epsilon and line width).
The user can then record this parameter set or cancel and try another fit pattern or some fit with different baseline range.

WARNINGS : If you find some obviously out of purpose fit, it likely for one of the three following reasons :


The selection is not large (short) enough. It is important to select enough but not too much baseline, particularly when baseline is highly increasing (or decreasing) and for absorption lines :
The first guess of fitting routines suppose a line located one the highest or lowest value of the selection.


Wrong pattern . If the pattern you choose does not correspond to your line you may have a fit completely different from the signal, for instance sinus pattern is often much too thin for LWS lines.


Too noisy lines . For the moment, nothing limits variation range of line parameters (for instance line width), for that reason too noisy lines are not fitted correctly.

Line database :

• a list of atomic infrared lines created by Karel van der Hucht ("SWS" line list)
• a list by Moore, "Identification List of Lines in Stellar Spectra"
• a list of 'all' SWS and LWS lines put by observers in ISO Central Programme observations.
• a list of molecular lines by J. Crovisier

• show recorded lines : display a window with the list of lines and their parameters
• print recorded lines : print of a file the line list (same format and as show button)
• clean recorded lines : erase all recorded informations
  

• Replot lines and fit :
  replot whole spectrum, overplot all recorded line fits, all baseline fits and write upper the first possible element recorded.
• Replot original data :
  Replot original spectrum without any overplot.
• Hardcopy lines and fit :
  Plot on a postscript file the whole spectrum, overplot all recorded line fits, all baseline fits and write upper the first possible element recorded.
• Hardcopy with lines only :
  Plot on a postscript file the whole spectrum, and write upper each recorded line, the first possible element recorded, no overplot.