The fit1d command

The fit1d command fits profiles along a particular axis of a NDF data file. It is a generic command that will work on hypercubes with upto 7 dimensions, but is here discussed in terms of a typical 3-D ACSIS data-file with axes RA, Dec, and Vlsr. Specifically, the fitting of spectra across the imaged section of the sky. The output of fit1d is a cube with the fitted profiles and ``Component parameter files as NDF extentions''. Be aware that the input cube is expected to be baseline subtracted and to have a zerolevel of 0.

What sets fit1d apart from most other fitting routines is that, by using ``Component parameter files'' as inputs, it gives individual control over the fitting at each position. For instance, it is possible to fit broad lines on the nucleus of a galaxy but narrow lines everywhere else in the disk. Or to fit multiple components in an outflow and single components everywhere else in the field. Still, these types of fits may require a considerable familiarity with handling, cutting, and pasting NDF files in order to ``create'' the desired parameter files for input.

Figure: Fit1d $-$ Black: original profiles; Red: results of a 3-component Gauss-Hermite2 fit (fitting both h3 and h4, see next section)

fit1d can also fit more complicated shapes than Gaussians. In particular, Gauss-Hermite functions are a powerful extention when fitting profiles that are skewed, peaky, or only approximately Gaussian. Figure shows Gauss-Hermite profiles as a function of the ``skewness'' coefficient h3 and the ``peakiness'' coefficient h4. The red box indicates the limits on acceptable values for h3 and h4 as defined in the defaults config file. The limits were chosen such as to exclude fits that look more like multiple components rather than a distorted single Gaussian, but, admittedly are fairly arbitrary.

Because of the ability to fit distorted shapes, Gauss-Hermites are particularly well suited to ``capture'' the maximum amount of emission from a cube. Figure shows an example of the quality of the fits that can be obtained. For the shown case fit1d used a 3-component gauss-hermite2 (fitting h3 and h4) function with the range around the profiles and the remaining configuration parameters at their default setting. Collapsing the cube with the fitted profiles can thus result in an accurate and almost noise-free white-light or total-emission map. Residuals from the fit can of course be studied by subtracting the fitted profiles from the original cube.