simonsobs / sotodlib

This was an issue during to SAT1 cooldown in the highbay. Most of the pieces were set up but we were missing at least one thing. So the data files are written to have high-precision timing but the data between the different UFMs were not actually co-sampled. Book binder needs to throw errors if this happens.

Cosampling check isn't done until binding step, should be done earlier in the bookbinder preprocess step here:

sotodlib/sotodlib/io/bookbinder.py

Lines 668 to 673 in da3e6e5

    
           t0 = np.max([s.times[0] for s in self.streams.values()]) 
        
           t1 = np.min([s.times[-1] for s in self.streams.values()]) 
        
           # prioritizes the last stream 
        
           # implicitly assumes co-sampled (this is where we could throw errors 
        
           # after looking for co-sampled data) 
        
           ts, _ = fill_time_gaps(stream.times)

	t0 = np.max([s.times[0] for s in self.streams.values()])
	t1 = np.min([s.times[-1] for s in self.streams.values()])
	# prioritizes the last stream
	# implicitly assumes co-sampled (this is where we could throw errors
	# after looking for co-sampled data)
	ts, _ = fill_time_gaps(stream.times)

Imprinter / Bookbinder co-sampling check