This fixes bugs related to how spikeinterface handles loading different blocks of recordings in read_openephys. Previously it was based on the experiment number, which is not always reliable (e.g. when some experiments are deleted). Now it is switched to based on the block_index as returned by the Neo package. It is much more reliable.
This fixes bugs related to how spikeinterface handles loading different blocks of recordings in
read_openephys. Previously it was based on the experiment number, which is not always reliable (e.g. when some experiments are deleted). Now it is switched to based on the block_index as returned by the Neo package. It is much more reliable.