Lesson 03: Publication Style¶

A consistent style across all figures in a paper signals professionalism and makes revision much easier. Changing the font size across fifteen figures is one edit instead of fifteen. matplotlib style sheets let you encode all style choices once and apply them everywhere.

This lesson covers the Nestling style file, LaTeX labels, and how to size figures correctly for journal submission.

---

The Nestling matplotlib style file¶

The style file lives at examples/plotting/styles/nestling.mplstyle. Its full contents and the rationale for each setting:

# Nestling matplotlib style
# Adapted from the Beacom-group conformal style.
# Optimised for publication-quality single or two-column figures.

text.usetex        : False
font.family        : serif
font.size          : 8.0

# Colour cycle — Dark2 palette (colourblind-friendly)
axes.prop_cycle    : cycler('color', ["#e7298a", "#1b9e77", "#d95f02", "#7570b3", "#66a61e", "#e6ab02"])

axes.grid          : False
legend.frameon     : False

axes.titlesize     : 8.0
axes.labelsize     : 8.0

xtick.direction    : in
ytick.direction    : in
xtick.top          : True
ytick.right        : True
xtick.labelsize    : 8
ytick.labelsize    : 8

figure.dpi         : 100
figure.figsize     : 3.0, 3.0

lines.linewidth    : 1.0

What each group does¶

Setting	Value	Why
text.usetex	False	No local LaTeX required; set to True to enable full LaTeX rendering (see below)
font.family	serif	Matches the body text in most physics journals
font.size	8.0	Legible at single-column print size without dominating the data
axes.prop_cycle	Dark2 colours	Colourblind-friendly; see Lesson 04
axes.grid	False	Gridlines add clutter; omit unless values need to be read off precisely
legend.frameon	False	The legend box wastes ink; remove it
xtick.direction / ytick.direction	in	Ticks point inward, conventional in HEP and astro
xtick.top / ytick.right	True	Mirror ticks for readability without a grid
figure.figsize	3.0, 3.0	Matches a single-column panel in most journals
lines.linewidth	1.0	Clean at print size; increase to 1.5 for slides

---

Applying the style¶

Load the style once at the top of your script:

import matplotlib.pyplot as plt
from pathlib import Path

STYLE = Path(__file__).parent / "styles" / "nestling.mplstyle"
plt.style.use(str(STYLE))

Every fig, ax = plt.subplots() call after this point uses the Nestling style.

You can also combine styles:

plt.style.use([str(STYLE), "path/to/your_overrides.mplstyle"])

Or override individual settings inline without a file:

import matplotlib as mpl
mpl.rcParams["lines.linewidth"] = 1.5   # heavier lines for slides

---

LaTeX labels¶

Without a local LaTeX installation¶

With text.usetex: False (the default), matplotlib uses its own math renderer. This covers the vast majority of physics notation:

ax.set_xlabel(r"$E$ / GeV")
ax.set_ylabel(r"$\phi_\nu$ / GeV$^{-1}$ cm$^{-2}$ s$^{-1}$ sr$^{-1}$")
ax.set_title(r"$\bar{\nu}_e$ flux")

Use raw strings (r"...") throughout to avoid having to escape backslashes.

With a local LaTeX installation¶

For full LaTeX rendering (real Computer Modern font, \text{}, custom \newcommands):

import matplotlib as mpl
mpl.rcParams["text.usetex"] = True
mpl.rcParams["text.latex.preamble"] = r"\usepackage{amsmath}"

Or add a one-line override style:

# tex_override.mplstyle
text.usetex: True

plt.style.use([str(STYLE), "tex_override.mplstyle"])

---

Figure sizing for journals¶

Most physics journals use one of two column widths:

Layout	Typical width
Single column	88 mm ≈ 3.46 in
Two-column (full page)	180 mm ≈ 7.09 in

The Nestling default (3.0 × 3.0 in) is close to single-column. Adjust for your target journal:

# Single column, 4:3 aspect
fig, ax = plt.subplots(figsize=(3.46, 2.6))

# Full width, two-panel
fig, axes = plt.subplots(1, 2, figsize=(7.09, 3.0))

---

Exporting¶

plt.savefig("figure.pdf", bbox_inches="tight")

• PDF: vector format, preferred by journals. Fonts are embedded.
• SVG: vector, editable in Inkscape. Useful for final touch-ups.
• EPS: accepted by many older journal workflows; avoid if PDF is accepted.
• PNG: raster only; use dpi=300 minimum. Set dpi=600 for figures with fine lines.

bbox_inches="tight" trims excess whitespace around the figure, which avoids awkward spacing when the figure is included in a LaTeX document.

---

A minimal complete example¶

from pathlib import Path
import matplotlib.pyplot as plt
import numpy as np

STYLE = Path("examples/plotting/styles/nestling.mplstyle")
plt.style.use(str(STYLE))

E = np.logspace(0, 6, 300)

fig, ax = plt.subplots()
ax.loglog(E, 1e-18 * (E / 1e5) ** -2.7, label=r"$E^{-2.7}$")
ax.loglog(E, 1e-18 * (E / 1e5) ** -2.0, label=r"$E^{-2.0}$")
ax.set_xlabel(r"$E$ / GeV")
ax.set_ylabel(r"$\phi$ / GeV$^{-1}$ cm$^{-2}$ s$^{-1}$ sr$^{-1}$")
ax.legend()
plt.tight_layout()
plt.savefig("spectrum.pdf", bbox_inches="tight")
plt.close(fig)

The full example is in examples/plotting/scripts/publication_plot.py.

---

What to read next¶

Lesson 04 covers colour in depth: why colourblind-friendly palettes matter, how to choose between sequential, diverging, and qualitative colour maps, and tools for checking your figures before submission.