Note
Go to the end to download the full example code
Handling probe information¶
In order to properly spike sort, you may need to load information related to the probe you are using.
SpikeInterface internally uses ProbeInterface to handle probes or probe groups for recordings.
Depending on the dataset, the Probe
object may already be included or might need to be set
manually.
Here’s how!
import numpy as np
import spikeinterface.extractors as se
First, let’s create a toy example:
recording, sorting_true = se.toy_example(duration=10, num_channels=32, seed=0, num_segments=2)
print(recording)
InjectTemplatesRecording: 32 channels - 30.0kHz - 2 segments - 600,000 samples - 20.00s
float32 dtype - 73.24 MiB
Segments:
Samples: 300,000 | 300,000
Durations: 10.00s | 10.00s
Memory: 36.62 MiB | 36.62 MiB
This generator already contains a probe object that you can retrieve directly and plot:
probe = recording.get_probe()
print(probe)
from probeinterface.plotting import plot_probe
plot_probe(probe)
Probe - 32ch - 1shanks
(<matplotlib.collections.PolyCollection object at 0x7efcecfd9270>, <matplotlib.collections.PolyCollection object at 0x7efd043e23e0>)
You can also overwrite the probe. In this case you need to manually set the wiring (e.g. virtually connect each electrode to the recording device). Let’s use a probe from Cambridge Neurotech with 32 channels:
from probeinterface import get_probe
other_probe = get_probe(manufacturer="cambridgeneurotech", probe_name="ASSY-37-E-1")
print(other_probe)
other_probe.set_device_channel_indices(np.arange(32))
recording_2_shanks = recording.set_probe(other_probe, group_mode="by_shank")
plot_probe(recording_2_shanks.get_probe())
ASSY-37-E-1 - cambridgeneurotech - 32ch - 2shanks
(<matplotlib.collections.PolyCollection object at 0x7efced44f850>, <matplotlib.collections.PolyCollection object at 0x7efcecbca260>)
Now let’s check what we have loaded. The group_mode=’by_shank’ automatically sets the ‘group’ property depending on the shank id. We can use this information to split the recording into two sub-recordings:
print(recording_2_shanks)
print(recording_2_shanks.get_property("group"))
rec0, rec1 = recording_2_shanks.split_by(property="group")
print(rec0)
print(rec1)
InjectTemplatesRecording: 32 channels - 30.0kHz - 2 segments - 600,000 samples - 20.00s
float32 dtype - 73.24 MiB
Segments:
Samples: 300,000 | 300,000
Durations: 10.00s | 10.00s
Memory: 36.62 MiB | 36.62 MiB
[0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1]
0
1
Note that some formats (MEArec, SpikeGLX) automatically handle the probe
geometry. For almost all other formats the probe and the wiring have
to be set manually using the probeinterface
library.
Total running time of the script: (0 minutes 0.295 seconds)