AgentSkillsCN

scientific-figure-pro

使用Python/matplotlib以统一的版式风格（类似Helvetica/Arial的无衬线字体、高对比度配色方案、极简的轴线设计、无边框图例，以及紧凑的导出默认设置）生成适合发表的科学图表。在需要创建条形图、趋势图、散点图，或多面板学术图表时使用此功能，这些图表必须符合存储库风格的审美要求，并满足可重复性约束。

SKILL.md
--- frontmatter
name: scientific-figure-pro
description: Generate publication-ready scientific figures in Python/matplotlib with a consistent house style (Helvetica/Arial-like sans fonts, high-contrast palettes, minimal spines, frameless legends, and tight export defaults). Use when creating bar, trend, scatter, or multi-panel academic figures that must match repository-style aesthetics and reproducibility constraints.

Scientific Figure Pro

Use the helpers in scripts/scientific_figure_pro.py to keep figure style consistent across tasks.

Apply Style

  1. •Import and call apply_publication_style(...) first.
  2. •Choose scale:
  • •Use FigureStyle(font_size=24, axes_linewidth=3) for dense bar/comparison panels.
  • •Use FigureStyle(font_size=15 or 16, axes_linewidth=2) for compact trend/scatter plots.
  1. •Keep text.usetex=False unless math typography is required and TeX is installed.

Use High-Level Plot Functions

  • •make_grouped_bar(ax, categories, series, labels, ...)
  • •annotate_bars(ax, bars, fontsize=12): Add exact values above bars.
  • •make_trend(ax, x, y_series, labels, ...)
  • •make_scatter(ax, x, y, groups=..., labels=...)
  • •make_sphere_illustration(ax, light_dir=...): Create a shaded 3D-effect sphere.
  • •finalize_figure(fig, out_path, dpi=300, pad=2)

These encode the house defaults:

  • •Ultra-Wide Scaling: Use wide figsize (e.g., width 3-4x height) for multi-metric panels.
  • •Legend Isolation: Consider dedicating the last axis in a grid to a shared legend (ax.set_axis_off()).
  • •Precision Ticks: Use FixedLocator for clean, whole-number y-ticks.
  • •Top/right spines removed
  • •Frameless legends
  • •Explicit y-limits with modest headroom
  • •Black bar edges for print-safe contrast
  • •Repository-aligned color palette

Reproducible Export Rules

  1. •Default to dpi=300.
  2. •Use dpi=600 for dense bar-heavy panels with small text.
  3. •Always call finalize_figure(...) to enforce tight_layout(pad=2) before saving.

Palette Policy

Use semantic colors from PALETTE:

  • •Blues (blue_main, blue_secondary) for key/proposed methods.
  • •Greens (green_1..green_3) for positive/improvement variants.
  • •Reds (red_1, red_2, red_strong) for contrasting baselines/ablations.
  • •neutral for references/background categories.
  • •highlight only for targeted callouts.

Quick Start

python
import importlib.util
from pathlib import Path
import sys

module_path = Path("skills/scientific-figure-pro/scripts/scientific_figure_pro.py")
spec = importlib.util.spec_from_file_location("scientific_figure_pro", module_path)
mod = importlib.util.module_from_spec(spec)
sys.modules[spec.name] = mod
spec.loader.exec_module(mod)

mod.apply_publication_style(mod.FigureStyle(font_size=16, axes_linewidth=2.5))
fig, ax = plt.subplots(figsize=(9, 5))
mod.make_grouped_bar(ax, categories, series, labels)
mod.finalize_figure(fig, "figure.png", dpi=300, pad=2)

If importing by package path is inconvenient, load the module by relative file path and reuse the same APIs.