Back in August 2018 we quickly put together the data writer, and just went with the first thing that could keep up with the data rate.
Ideally we'd figure out how to get FITS to keep up (for whatever reason it was very slow).
In the mean time, we may want to switch to something like hdf5 because npz's are not portable.
Back in August 2018 we quickly put together the data writer, and just went with the first thing that could keep up with the data rate. Ideally we'd figure out how to get FITS to keep up (for whatever reason it was very slow). In the mean time, we may want to switch to something like hdf5 because npz's are not portable.