Open PicniCat opened 1 year ago
Hi. thanks for the feeback. Yes by design sortingcomponents return low level buffer generaly numpy array or numpy with multi fields.
All the end user machinery on top on sorting and waveformextractor do not work here. peaks are not spike yet because they do not have "label"
If you consider the channel as the label in the peak vector you can do stuff like this:
sorting = NumpySorting.from_times_labels(peaks['sample_ind'], peak['channel_ind'], sampling_rate)
But we plan to make some widgets on top of this sortingcomponents vector but at the moment the best to make standard matplotlib and check these examples https://spikeinterface.github.io/
Hello,
When I was using
spikeinterface.sortingcomponents
, it was really hard to visualize some of the outcomes. This is because the functions, likepeak_detection
orpeak_localization
only return an ndarray instead of a sorting object. This makes it hard to be used inspikeinterface.widgets
. Could you let some of the functions inspikeinterface.widgets
likesw.plot_rasters
to be compatible with ndarray objects instead of only sorting objects?Also, it is hard to extract waveforms from the outcome of
peak_detection
. Are there any methods to manually pack up the outcome ofspikeinterface.sortingcomponents
into a sorting object?Many thanks!