viziphant.patterns.plot_patterns_hypergraph

viziphant.patterns.plot_patterns_hypergraph(patterns, num_neurons=None)

Hypergraph visualization of spike patterns.

The spike patterns are interpreted as a hypergraph. Neurons are interpreted as vertices of the hypergraph while patterns are interpreted as hyperedges. Thus, every pattern connects multiple neurons.

Neurons are depicted as circles on a 2D diagram. A graph layout algorithm is applied to the hypergraph in order to determine suitable positions for the neurons. Neurons participating in common patterns are placed close to each other while neurons not sharing a common pattern are placed further apart.

Each pattern is drawn, based on this diagram of neurons, in such a way that it illustrates which neurons participated in the pattern. The method used for this is called the subset standard. The pattern is drawn as a smooth shape around all neurons that participated in it.

The shapes of the patterns are colored such that every pattern has its own color. This makes distinguishing between different patterns easier, especially if their drawings overlap.

Parameters:

patternsdict or list of dict

One or more patterns from a list of found patterns returned by elephant.spade.spade() or elephant.cell_assembly_detection.cell_assembly_detection() pattern detectors.

num_neurons: None or int

If None, only the neurons that are part of a pattern are shown. If an integer is passed, it identifies the total number of recorded neurons including non-pattern neurons to be additionally shown in the graph. Default: None

Returns:

A handle to a matplotlib figure containing the hypergraph.

Examples

Here, we show an example of plotting random patterns from the CAD method:

import matplotlib.pyplot as plt
import numpy as np
import quantities as pq
from elephant.cell_assembly_detection import cell_assembly_detection
from elephant.conversion import BinnedSpikeTrain
from elephant.spike_train_generation import compound_poisson_process
import viziphant

np.random.seed(30)
spiketrains = compound_poisson_process(rate=15 * pq.Hz,
    amplitude_distribution=[0, 0.95, 0, 0, 0, 0, 0.05], t_stop=5*pq.s)
bst = BinnedSpikeTrain(spiketrains, bin_size=10 * pq.ms)
bst.rescale('ms')
patterns = cell_assembly_detection(bst, max_lag=2)

fig = viziphant.patterns.plot_patterns_hypergraph(patterns)
plt.show()

(Source code)