uvfits reader assumes presence of non-standard uvfits keywords

[bhazelton]

The pyFHD uvfits reader assumes that obsra or ra and obsdec or dec are present in the uvfits header. These are completely nonstandard. They often exist in MWA data, but not always. Instead the RA/Dec should be extracted from the appropriate axis info (the RA axis is either axis 5 or 6 depending on whether there's a spectral window axis; the Dec axis is either 6 or 7). For an example of how to do this properly, see pyuvdata's uvfits reader.

[bhazelton]

It's also assuming that a telescope location will be specified with lat/lon/alt keywords in the uvfits header which is nonstandard. That information is stored in the antenna table. See pyuvdata for the proper implementation.