made some small tweaks to the setup.py script so that the files get explicit names. My version of wget was downloading these files as filefile.1file.2 etc.
I also made some tweaks to the shifting diagnostic plots.
blue to red goes in order of increasing wavelength.
I've notice that there are still some outliers in the shift values, and it probably makes sense to implement some iterative outlier rejection. I could image a scheme along the following lines:
Compute a robust fit to shift vs. pixel value
Compute a robust sigma about this best fit (e.g. median absolute deviation)
If a given segment yields a shift that's more than 2 sigma from the robust fit, replace it with the robust value
Fit each order independently using the cleaned shift vs. pixel values.
made some small tweaks to the setup.py script so that the files get explicit names. My version of wget was downloading these files as
filefile.1file.2etc.I also made some tweaks to the shifting diagnostic plots.
blue to red goes in order of increasing wavelength.
I've notice that there are still some outliers in the shift values, and it probably makes sense to implement some iterative outlier rejection. I could image a scheme along the following lines: