For the test with alternative fitting functions we selected the galaxy IC 3322A J, which showed in the 1D constrained two disk fit a good set of profiles. The best quadrant of the galaxy was the lower north-west quadrant, which was the one we used to compare to the alternative fitting functions. The results are shown in Table 10.
An attempt to do a Sérsic Law fit on IC 3322A resulted in parameters values we could not discuss or derive conclusions from. Since the Generalized Gaussian proved to provide acceptable parameter values and artifical galaxy tests showed no difference between the Generalized Gaussian and the Sérsic Law in the matter of fit quality, we decided to only use the Generalized Gaussian (see Section 4.8).
The results are shown in Table 10B. A plot of the two fits on the first profile is shown in figure 7 where the red line represents the 1D Constrained Two Disk fit and the blue line the Generalized Gaussian fit. For easy comparison the 1D Constrained Two Disk fit table of the same area has been added. The value for stays for a range of values rather constant, showing where the general good fit lies. This is also directly related to which behaves similarly. The values for show no coherency, but unlike the constrained two disk fit the inner part is not constrained, which is probably the cause for this. The main problem with the results from the Generalized Gaussian is the interpretation of the parameter values, as there is no trend visible and the parameters are more erratic. So our idea of using less parameters and create a more stable fit because less parameters are less sensitive to irregularities, did not work out
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Notes: (1)(6) See Section 3.6 for the defitinion. (2) central radial position of the bin. (3) Thin disk central surface brightness. (4) Thick disk central surface brightness. (5) Thick/thin disk scaleheight ratio. (7) Central surface brightness. (8) Width of the distribution. (9) Shape parameter. |