I think I have discovered an inconsistency in the d20200109_HIRANO_PDB dataset:
ds = csr_dataset("d20200109_HIRANO_PDB")
unique(ds$ports$CSR_PORT) # gives 0
unique(ds$data$CSR_PORT) # gives 1 2 3 4 5 6
To my mind, all ports should be referenced in ports, so that the user knows which CSR_MSMT_VAR they are looking at.
Thanks for putting this great effort together!