This release adds a fix for warnings related to conflicting BLAS and OpenMP implementations, as well as other misc bug fixes. It also adds logging for resource usage throughout the sorting pipeline.
This release adds: - Option to specify channels to exclude from sorting through the API and GUI. - Documentation for output files. - A new tutorial notebook demonstrating how to modify good/mua labels using custom criteria. - Option to force-clear PyTorch's reserved GPU memory between some operations, using `clear_cache = True`.
This release contains several bug fixes, changes the `duplicate_spike_bins` parameter to `duplicate_spike_ms` (new default 0.25), adds the `highpass_cutoff` parameter, and improves logging of error messages.
This release adds the option to save a drift-corrected copy of the data to `temp_wh.dat` in the results directory. This can be done through the API using `save_preprocessed_copy=True` for `run_kilosort`, or by clicking the checkbox with the same name under the "Run" button in the GUI. Using either option will also update options for Phy so that this copy of the data is used instead of the raw data.
In addition to some minor fixes and additions to logging behavior, this release addresses the following problems: * Negative spike times will no longer be detected. * Plots generated by GUI will be saved in the results directory in .png format. * .mat probe files will correctly load kcoords (shank index) for probes with disconnected channels. * template placement is handled separately for each shank, so that the spacing between shanks will not affect template positions. * added new shift and scale parameters for handling float32 data better.
This release adds a new parameter, `drift_smoothing`, which controls the amount of gaussian smoothing applied to the spatiotemporal drift estimation. It also switches control of console output to the logging module and adds a `kilosort.log` file to the sorting results.