def _calc_features(self):
spikes = self.spikes_src.spikes
feats = [f(spikes) for f in self.feature_methods.values()]
ft_data = features.combine(feats,
norm=self.normalize,
feat_method_names = self.feature_methods.keys())
# Filter feature_data to remove _hidden_features.
# This routine is O(n^2), to deal woth possible repetitions
names, idx = [], []
for i, name in enumerate(ft_data['names']):
if name not in self._hidden_features:
names.append(name)
idx.append(i)
ft_data['names'] = names
ft_data['data'] = ft_data['data'][:, idx]
self._feature_data = ft_data
def _calc_features(self):
spikes = self.spikes_src.spikes
feats = [f(spikes) for f in self.feature_methods.values()]
ft_data = features.combine(
feats,
norm=self.normalize,
feat_method_names=self.feature_methods.keys())
# Filter feature_data to remove _hidden_features.
# This routine is O(n^2), to deal woth possible repetitions
names, idx = [], []
for i, name in enumerate(ft_data['names']):
if name not in self._hidden_features:
names.append(name)
idx.append(i)
ft_data['names'] = names
ft_data['data'] = ft_data['data'][:, idx]
self._feature_data = ft_data
#!/usr/bin/env python
#coding=utf-8
from spike_sort.io.filters import PyTablesFilter
from spike_sort import extract
from spike_sort import features
from spike_sort import cluster
from spike_sort.ui import plotting
import os
dataset = '/SubjectA/session01/el1'
datapath = '../../../data/tutorial.h5'
io_filter = PyTablesFilter(datapath)
raw = io_filter.read_sp(dataset)
spt = extract.detect_spikes(raw, contact=3, thresh='auto')
sp_win = [-0.2, 0.8]
spt = extract.align_spikes(raw, spt, sp_win, type="max", resample=10)
sp_waves = extract.extract_spikes(raw, spt, sp_win)
sp_feats = features.combine(
(features.fetP2P(sp_waves), features.fetPCs(sp_waves)))
clust_idx = cluster.cluster("gmm", sp_feats, 4)
plotting.plot_features(sp_feats, clust_idx)
plotting.show()
io_filter.close()
#!/usr/bin/env python
# -*- coding: utf-8 -*-
from spike_sort.io.filters import PyTablesFilter
from spike_sort import extract
from spike_sort import features
from spike_sort import cluster
from spike_sort.ui import plotting
import os
dataset = '/SubjectA/session01/el1'
datapath = '../../../data/tutorial.h5'
io_filter = PyTablesFilter(datapath)
raw = io_filter.read_sp(dataset)
spt = extract.detect_spikes(raw, contact=3, thresh='auto')
sp_win = [-0.2, 0.8]
spt = extract.align_spikes(raw, spt, sp_win, type="max", resample=10)
sp_waves = extract.extract_spikes(raw, spt, sp_win)
sp_feats = features.combine(
(
features.fetP2P(sp_waves),
features.fetPCA(sp_waves)
)
)
plotting.plot_features(sp_feats)
plotting.show()
def _calc_features(self):
spikes = self.spikes_src.spikes
feats = [f(spikes) for f in self.feature_methods]
self._feature_data = features.combine(feats, norm=self.normalize)
def _calc_features(self):
spikes = self.spikes_src.spikes
feats = [f(spikes) for f in self.feature_methods]
self._feature_data = features.combine(feats, norm=self.normalize)