Python Script ~6 KB Free

Matplotlib Publication Figure Template

A Python script with pre-configured rcParams and helper functions that produce clean, publication-ready figures matching journal style guides — without wrestling with matplotlib defaults.

What's Included

rcParams preset for journal-quality output (300 DPI, correct font sizes)
Single-column and double-column figure size helpers
Clean colour palettes (colour-blind friendly)
Helper functions: annotate_significance(), add_panel_labels(), save_fig()
Line plot, bar chart, scatter plot, and boxplot examples
LaTeX-compatible font rendering settings
requirements.txt with matplotlib, numpy, scipy

Who This Is For

Researchers submitting to journals with strict figure formatting requirements
PhD students producing thesis figures that look professional
Anyone tired of reformatting matplotlib output before submission

What Makes a Publication-Quality Figure

Journal requirements typically demand:

300 DPI minimum (often 600 for line art)
Figure width: 86 mm (single column) or 176 mm (double column) — exact values depend on journal
Font size: 8–10pt for axis labels, consistent with body text
No chart junk: white background, no box spines, minimal grid lines
Colour-blind friendly palettes (avoid red/green alone)

This template sets all of these automatically. A single import figure_styles at the top of a notebook applies the preset.

Example Usage

import matplotlib.pyplot as plt
import numpy as np
from figure_styles import apply_journal_style, save_fig, single_col_width

apply_journal_style()

fig, ax = plt.subplots(figsize=(single_col_width, single_col_width * 0.75))
ax.plot(x, y, color='#2563eb', lw=1.5, label='Treatment')
ax.set_xlabel('Time (days)')
ax.set_ylabel('Response (AU)')
ax.legend(frameon=False)

save_fig(fig, 'my_figure', formats=['pdf', 'svg'])

Output: my_figure.pdf and my_figure.svg at the correct DPI, ready for submission.

Ready to get started?

Enter your email and I'll send the link straight to your inbox.

Prefer to browse the source? View on GitHub →