Make quick invert rest wavelength calculation the same as in dynamics

[mgalloy]

The mask for the pixels to use in the rest wavelength calculation are not the same for the dynamics and quick invert files. The quick invert calculation should be changed to be the same as the dynamics calculation.

[mgalloy]

The rest wavelength should be computed from the uncorrected velocity correct? For example, it is dop, not corrected_dop, in the following calculation:

east_median_rest_wavelength = median(dop[good_east_dop_ind])

The corrected velocity, corrected_dop, is only used again for the values in the "Doppler Velocit" extension.

I think this was a bug in the quick invert rest wavelength calculation.

[detoma]

We are computing the rest wavelength so we can create a corrected_dop. The rest wavelength should be subtracted from the uncorrected dop.

[mgalloy]

I think this is done, will see from test in #135.

[detoma]

Make sure the old code comp_doppler_correction.pro is not used any longer, or, if used, is updated and include the new rest wavelength determination.

Make sure the rest wavelength is not subtracted from velocities that are equal to zero so we do not create erroneous velocity values.