Open dkirkby opened 8 years ago
Currently, EarthLocation.of_site() requires an active internet connection to look up APO.
It looks like the download only happens the first time and then is cached:
https://github.com/astropy/astropy/blob/v1.1.x/astropy/coordinates/earth.py#L258
Hmm, I can't reproduce the error now but this was the error I encountered a few days ago:
WARNING: Could not access the online site list. Falling back on the built-in version, which is rather limited. If you want to retry the download, do EarthLocation._get_site_registry(force_download=True) [astropy.coordinates.earth]
Traceback (most recent call last):
...
self.where = astropy.coordinates.EarthLocation.of_site('apo')
File "/Users/Daniel/anaconda/lib/python2.7/site-packages/astropy/coordinates/earth.py", line 220, in of_site
raise UnknownSiteException(e.site, 'EarthLocation.get_site_names', close_names=e.close_names)
astropy.coordinates.errors.UnknownSiteException: u"Site 'apo' not in database. Use EarthLocation.get_site_names to see available sites."
The
specsim.transform.observatories
dictionary gives astropy coordinates of APO, KPNO, LSST by name, as a convenience. This issue is to verify that these are correct and investigate if there is an alternative source of this info that should be used instead. This was prompted by my recent discovery that astropy has an extensive database that is accessible via EarthLocation.of_site() and stored in a json file.However, taking the case of APO, it is not clear what the right answer should be since there are two different telescopes (2.5m and 3.5m). Are the differences between their locations negligible for all applications? The APO home page quotes:
while the astropy db uses (from the IRAF Observatory Database):
and Don's nightly science plans assume:
Finally, the (e)BOSS plate design code uses: