We read every piece of feedback, and take your input very seriously.
To see all available qualifiers, see our documentation.
Updated Kwik format (markdown)
Fixed cluster_groups.
Having shank numbers as strings in the channel_groups dictionary in the .prb file causes key error.
Added quick notes on MATLAB.
Added `channel_order` field, specifying the order of the channels in the PRB file.
Merge branch 'master' of https://github.com/klusta-team/kwiklib.wiki
Updated spec of the Kwik format: there is no longer a channel_map field. The mapping of channel indices between the raw data array and the probe file is now necessarily trivial.
Got rid of IGNORED_CHANNELS, it is now in the PRB file.
In the PRB file, channel_groups is now a dict instead of a list.
Added "id" to channel_group in PRB file
Renamed nfeatures per channel.
Renamed sample rate
Added precisions on the data type of a few attributes.
Updated kwik format: added channel_map, and the channel indices are now relative instead of absolute (i.e. one must specify the channel_group too).
Support for multiple clusterings.
Updated probe file (now Python).
Added metadata in .kwd files and note about open-ephys.
There is now a "main" clustering. The "clusters" group refers to this clustering. Additional clusterings are read-only and do not come with metadata.
Minor changes to the format specification. voltage_gain is in microvolts kwik_version in all HDF5 files clusters is now a group with at least to datasets: original and manual. Other clusterings may be added here.
Updated the file format spec to the last version (in spikedetekt2/dev repo).
Added description of KWE file with events.
Added meeting notes.
Updated the file format specifications with latest decisions.
Added notes about e-mail discussions late August.