ColorbarSpec

Describes the colorbar attached to plot_image and plot_hexbin. You rarely build one by hand thanks to the colorbar= argument.

colorbar= accepts True (default bar), a str (its label), or a ColorbarSpec.

import numpy as np
import behaviz as bv
import matplotlib.pyplot as plt

fig,axs = plt.subplots(1,3,figsize=(18,6,)) # for matplotlib and seaborn backends

data = np.random.default_rng(0).normal(size=(40, 60))
fig, ax = bv.plot_image(data, cmap="magma")

bv.plot_image(matrix, ax=ax[0], cmap="magma", colorbar=True)        # default bar
bv.plot_image(matrix, ax=ax[1], cmap="magma", colorbar="Correlation")  # str → labelled bar
bv.plot_image(matrix, ax=ax[2], cmap="magma",
              colorbar=bv.ColorbarSpec(label="r", location="bottom", fontsize=14))  # full control

Fields

Field	Default	Meaning
label	""	colorbar label
location	"right"	right / left / top / bottom
ticks	None	explicit tick positions
tick_fmt	None	printf format, e.g. "%.1f"
fraction	0.046	bar size as a fraction of the axes (matches axes height by default)
pad	None	gap between axes and bar (None → sensible default)
fontsize	12	label + tick-label font size

matplotlib / seaborn

Colorbars are a matplotlib construct; behaviz renders them on the matplotlib and seaborn backends. The bokeh backend draws the image itself but does not yet attach a behaviz colorbar.