uvfits key describing the number of timesteps

[cjordan]

Not sure if there's an official uvfits key or not, but it's quite painful discovering the number of timesteps in a uvfits file -- one more or less has to read the entire file to determine it. A key (even a non-standard one) would be very useful. I guess using this key would also assume that the layout of the uvfits is regular (i.e. one timestep has x rows/baselines before the next timestep which also has x rows, etc.).

[d3v-null]

Yeah, I would probably call this NSCANS to be consistent with metafits and mwaf

[cjordan]

I just realised that this doesn't reveal what times are actually available in a uvfits file. What do you think about providing unique timestamps with some key too?