Plotting multiple sequences

This snippet shows you how to plot a number of sequences as one combined event plot.

Simple example

Let’s first start by importing what we are going to need:

from thebeat.core import Sequence, SoundStimulus, SoundSequence
from thebeat.visualization import plot_multiple_sequences
import numpy as np

# Additionally suppress warnings to avoid confusion
import warnings
warnings.filterwarnings('ignore')
%matplotlib inline

---

We start by creating two dummy sequences:

# This is for reproducible randomness (so you will get the same sequences as me)
rng = np.random.default_rng(seed=123)

# Create sequences
seq_isoc = Sequence.generate_isochronous(n_events=8, ioi=500)
seq_random = Sequence.generate_random_normal(n_events=8, mu=500, sigma=100, rng=rng)

Now we make a list of these sequences, and plot using plot_multiple_sequences().

seqs = [seq_isoc, seq_random]
plot_multiple_sequences(seqs);

Complex example

In this example we generate 10 sequences with random inter-onset intervals (IOIs), each containing 5 events that also have random event durations.

# For reproducible randomness
rng = np.random.default_rng(seed=123)

# Generate random sequences using
seqs = [Sequence.generate_random_uniform(n_events=5, a=200, b=800) for _ in range(10)]

# Create empty list that we will fill with SoundSequence objects
stimseqs = []

# Loop over the sequences
for seq in seqs:
    # Generate 5 events with random durations
    stims = [SoundStimulus.generate(duration_ms=x) for x in rng.integers(low=10, high=100, size=5)]
    # Create a SoundSequence and add to stimseqs list
    stimseqs.append(SoundSequence(stims, seq))

# Additionally give them names
sequence_names = [f"Random {i+1}" for i in range(len(stimseqs))]

# Now plot!!
plot_multiple_sequences(stimseqs, y_axis_labels=sequence_names);

Adjusting the figure

Now, of course the standard settings for the plot are great, but what if we want to adjust the plot some more?

The plotting functions and methods used above all return a matplotlib Figure object, and a matplotlib Axes object.

These objects we can manipulate, and then show or save, as in the example below.

import matplotlib.pyplot as plt

# Generate random sequences (use a generator with a seed so you get the same results as we)
rng = np.random.default_rng(seed=123)
seqs = [Sequence.generate_random_normal(n_events=10, mu=500, sigma=50, rng=rng) for _ in range(10)]

# Use the Sequence.plot() method
fig, ax = plot_multiple_sequences(seqs,
                                  title='My random sequences')

# Add in some text with a box around it
box_properties = dict(boxstyle='round', facecolor='white', alpha=0.8)
ax.text(4300, 0.01, s="$\mu = 500$\n$\sigma = 50$", bbox=box_properties, fontsize=14)

# Show
fig.show();