Reading and decoding the data stored in baseband files can be somewhat slow, especially if the analysis itself is simple. So far, code development has focussed more on correctness than on performance, but a few things can help.
If you have other tips on performance or contributions that help improve it, please raise an issue or make a pull request!
Once you know a file does not have missing frames or is otherwise
slighly corrupted, you can speed up reading by turning off
verification of headers, by passing in
verify=False when opening
the stream reader.
Like python file readers in general, baseband’s stream readers cannot
be used in parallel threads. They can, however, be sent from process
to process using pickle
(or copied using