kis_tools issueshttps://gitlab.leibniz-kis.de/sdc/kis_tools/-/issues2021-05-06T14:48:42Zhttps://gitlab.leibniz-kis.de/sdc/kis_tools/-/issues/282wcs wavelength retrieval shifted by one pixel2021-05-06T14:48:42ZCarl Schafferwcs wavelength retrieval shifted by one pixelPhilip pointed out to me that the wcs retrieval of wavelengths might be off by one pixel. We narrowed the problem down to the definition of the `CRPIX`keyword. We set it to `0` meaning to indicate the beginning of the array, but using th...Philip pointed out to me that the wcs retrieval of wavelengths might be off by one pixel. We narrowed the problem down to the definition of the `CRPIX`keyword. We set it to `0` meaning to indicate the beginning of the array, but using the `WCS.wcs_pix2world(pixels,0)` method to retrieve the value yields a wavelength array with larger by `CDELT`than we would expect.
We either need to set the CRPIX keyword to 1 or use 1 for the origin value of the retrieval method.
It is unclear which of these solutions should be preferred.
Links:
* [related issue on github](https://github.com/astropy/astropy/issues/9989)
* [wcs paper](https://www.aanda.org/articles/aa/pdf/2002/45/aah3859.pdf)
* [astropy docs for WCS.wcs_pix2world](https://docs.astropy.org/en/stable/api/astropy.wcs.WCS.html#astropy.wcs.WCS.wcs_pix2world)Carl SchafferCarl Schaffer