Crack SpyKing PC Spy 5 0
Teodolinda Mattson
Note that filtering is performed only once, and if the code is relaunched on the same data, a flag in the parameter file will prevent the code to filter twice. You can specify only a subset of steps by doing:
The code can accept a text file with several commands that will be executed one after the other, in a batch mode. This is interesting for processing several datasets in a row. An example of such a text file commands.txt would simply be:
To be sure that data are properly loaded before filtering everything on site, the code will load only the first second of the data, computes thresholds, and show you an interactive GUI to visualize everything. Please see the documentation on Python GUI
Launch an interactive GUI to show you, superimposed, the activity on your electrodes and the reconstruction provided by the software. This has to be used as a sanity check. Please see the documentation on Python GUI
If the preview mode is activated, by default, it will show the first 2 seconds of the data. But you can specify an offset, in second, with this extra parameter such that the preview mode will display the signal in [second, second+2]
If you want to generate synthetic benchmarks from a dataset that you have already sorted, this allows you, using the benchmarking mode, to produce a new file output based on what type of benchmarks you want to do (see type)
While generating synthetic datasets, you have to chose from one of those three possibilities: fitting, clustering, synchrony. To know more about what those benchmarks are, see the documentation on extra steps
The code, when launched for the first time, generates a parameter file. The default template used for the parameter files is the one located in /home/user/spyking-circus/config.params. You can edit it in advance if you are always using the same setup.
Spike sorting is comprised of several steps, or components. In the spikeinterface.sortingcomponents module weare building a library of methods and steps that can be assembled to build full spike sorting pipelines.
However, this might not be the best option. It is in fact very likely that a sorter has one excellent step,say the clustering, but another step, which is sub-optimal. Decoupling different steps as separate components would allowone to mix-and-match sorting steps from different sorters.
Another advantage of modularization is that we can accurately benchmark every step of a spike sorting pipeline.For example, what is the performance of peak detection method 1 or 2, provided that the rest of the pipeline is thesame?
When too many peaks are detected a strategy can be used to select (or sub-sample) only some of them before clustering.This is the strategy used by spyking-circus and tridesclous, for instance.Then, clustering is run on this subset of peaks, templates are extracted, and a template-matching step is run to findall spikes.
Several methods have been proposed to correct for drift, but only one is currently implemented in SpikeInterface.See Decentralized Motion Inference and Registration of Neuropixel Datafor more details.
border_mode is a very important parameter. It controls dealing with the border because motion causes units on theborder to not be present throughout the entire recording. We highly recommend the border_mode='remove_channels'because this removes channels on the border that will be impacted by drift. Of course the larger the motion isthe greater the number of channels that would be removed.
7fc3f7cf58