Suggestion: Sigma Clipping, automated restart of stacking, Zoom

[TheSkyseeker]

Hi Guys, Filtering outliers via Kappa-Sigma Clipping would be very useful. For reducing impact on performance, a choosable binning may be interesting (for those with bigger sensors but doing EAA)

Also an automated restart of stacking if slewing to a new target would be useful. Ether by detecting repeated alignment failures to assume user slewed to a new target, or a threshold if the stack changes greater than e.g. 10%

Using ALS remotely (VNC..) via Tablet-PC makes it sometimes difficult to use drag-gestures for zooming in and out. What about a 100% and a fit to screen Button on the right panel above the processing block.

Last point, I do not know how intesitiv a FWHM calculation is, but does it, all things considered, reduce stacking time when discarding bad frames?

Greetings Steven

[gehelem]

Thank you for your feedback Another idea, just to mention it : Target changes could be detected by checking FITS headers, wich (often?) contain position information

[gnthibault]

Example of how to implement that with scipy/numpy for 5 sigma symmetric (I have used that used in another soft):

        import numpy as np
        import scipy.stats as scs
        mean = np.apply_along_axis(lambda x:
            np.mean(scs.sigmaclip(x, low=5.0, high=5.0)[0],
                    dtype=np.float32),
            2, stack)