AxisSpec

spec.x and spec.y are each an AxisSpec — independent, so the two axes can differ in scale, ticks, grid, spines, everything below.

import behaviz as bv
import numpy as np
import polars as pl

x = np.linspace(0, 2 * np.pi, 100)
y = np.sin(x)

spec = bv.PlotSpec(
    x=bv.AxisSpec(label="Time", unit="s", scale=bv.ScaleType.LOG),
    y=bv.AxisSpec(label="Voltage", unit="mV", scale=bv.ScaleType.LINEAR),
)

bv.plot_line("t", "v", data=df, spec=spec)

Cross-backend

Every field below takes effect on matplotlib, seaborn and bokeh, except the two marked matplotlib only (spine_offset, spine_trim) — bokeh has no spine model — and tick_sides, which on bokeh can only toggle the primary side (bottom/left).

Fields

Field	Default	Meaning
label	""	axis label
unit	""	appended automatically → "Voltage (mV)"
fontsize	12	label + tick-label font size
scale	"linear"	linear / log / symlog / logit
lim	None	(min, max) or None → auto
ticks	None	explicit tick positions (numbers, or strings → categorical labels)
tick_fmt	None	printf format, e.g. "%.2f"
invert	False	flip axis direction
spines	all four	which spines to draw
spine_width	2	spine line width
spine_color	None	spine colour (None → backend default)
spine_offset	0	push spine outward, px — matplotlib only
spine_trim	False	clip spine to the outer ticks — matplotlib only
tick_dir	"out"	tick direction: out / in / inout
tick_length	None	tick length (None → 3 × spine_width)
tick_width	None	tick line width (None → spine_width)
tick_color	None	tick colour
tick_sides	None	sides that show tick marks (["bottom"], ["left","right"], …)
grid	True	major grid on
grid_minor	False	minor grid on
grid_color	"#c1c1c1"	grid colour
grid_alpha	0.5	grid opacity
grid_style	"-"	major grid linestyle (-, --, :, -.)
grid_width	0.8	major grid line width

Examples

Labels, units, scales, limits font size... everything!

spec = bv.PlotSpec(
    x=bv.AxisSpec(label="t", fontsize=16, lim=(0, 10)),
    y=bv.AxisSpec(unit="Hz", fontsize=16, lim=(1e-3, 1e2), scale=bv.ScaleType.LOG, grid_minor=True),
)

bv.plot_line("t", "v",data=df, spec=spec)

Ticks: direction, length, width, colour

spec = bv.PlotSpec(
    x=bv.AxisSpec(label="t",fontsize=16,lim=(0, 10)),
    y=bv.AxisSpec(tick_dir="in", tick_length=8, tick_width=2, tick_color="#E023E0"),
)

bv.plot_line("t", "v", ax=ax,data=df, spec=spec)

Spines: subset, colour, despine offset/trim

# only the left & bottom spines, pushed 8px outward and trimmed to the data
spec = bv.PlotSpec(
    x=bv.AxisSpec(label="t",fontsize=16,lim=(0, 10)),
    y=bv.AxisSpec(spines=["left", "bottom"], spine_color="#EC2525",spine_offset=8, spine_trim=True),
) # offset/trim: matplotlib/seaborn

bv.plot_line("t", "v", ax=ax,data=df, spec=spec)  

Same call, three backends

A single spec renders the same on every backend:

import numpy as np
import polars as pl
import behaviz as bv
from bokeh.io import show, output_notebook

x = np.linspace(0, 2 * np.pi, 100)
y = np.sin(x)

df = pl.DataFrame({"t":x,"v":y})

spec = bv.PlotSpec(
        x=bv.AxisSpec(label="t", unit="s", spine_color="#EC2525",tick_dir="in"),
        y=bv.AxisSpec(label="signal", spines=["left", "bottom"], grid_width=2, grid_style=":",grid_color="#0010a4"),
    )

matplotlibseabornbokeh

bv.set_renderer("matplotlib")
fig, ax = bv.plot_line("t", "v", data=df, spec=spec)

bv.set_renderer("seaborn")
fig, ax = bv.plot_line("t", "v", data=df, spec=spec)

bv.set_renderer("bokeh")
fig, ax = bv.plot_line("t", "v", data=df, spec=spec)
show(fig)