科研 Skill

材料科学图表绘制规范

materials-figures

Covers publication-quality figure conventions for materials science plots: phase diagrams, pole figures, stress-strain curves. Use when the deliverable is a figure, not just an analysis result.

发布者 Codar安全审核通过v0.1.0更新于 2026-11-18

SKILL.md

Materials Science Figures

Overview

A correct analysis can still be misrepresented by a misleading figure — this skill covers the conventions specific to common materials-science plot types that make the difference between a figure that accurately conveys uncertainty/methodology and one that doesn't, on top of general good plotting practice.

When to use this skill

Activate when the request involves:

  • Making a publication figure or plot from materials characterization or simulation results
  • Terms: phase diagram plot, pole figure, stress-strain curve figure, band structure plot, DOS plot
  • "Make a figure of this result," "plot this phase diagram," "visualize this stress-strain data"

Core usage

Stress-strain curve with clearly marked derived quantities

import matplotlib.pyplot as plt

fig, ax = plt.subplots()
ax.plot(strain, stress, color="black")
ax.axhline(yield_strength, color="gray", linestyle="--", linewidth=0.8, label=f"Yield (0.2% offset): {yield_strength:.0f} MPa")
ax.scatter([strain[np.argmax(stress)]], [stress.max()], color="red", zorder=5, label=f"UTS: {stress.max():.0f} MPa")
ax.set_xlabel("Engineering strain")   # state explicitly: engineering, not true strain — see mechanical-testing pitfalls
ax.set_ylabel("Engineering stress (MPa)")
ax.legend()

Phase diagram (pymatgen's plotter, with stability annotation)

from pymatgen.analysis.phase_diagram import PDPlotter

plotter = PDPlotter(phase_diagram, show_unstable=0.05)   # show near-hull unstable entries up to 50 meV/atom
plot = plotter.get_plot()
# Label energy-above-hull explicitly for any entry called out in the figure —
# a phase diagram figure without hull-distance context invites readers to
# assume everything shown is equally stable, which is rarely the intent.

Pole figures (crystallographic texture)

# Pole figures plot crystallographic orientation density on a stereographic
# projection — typically produced by dedicated texture-analysis software
# (e.g. MTEX, a MATLAB toolbox) rather than general plotting libraries, since
# the stereographic projection and orientation-distribution-function machinery
# is specialized. When building one manually, always include the projection
# convention (upper/lower hemisphere) and intensity scale explicitly.

Band structure and density of states (combined plot)

from pymatgen.electronic_structure.plotter import BSDOSPlotter

plotter = BSDOSPlotter(bs_projection="elements", dos_projection="elements")
plot = plotter.get_plot(band_structure, dos)
# Mark the Fermi level explicitly (usually a horizontal line at E=0 by
# convention) — a band structure/DOS plot without a clear Fermi level reference
# is not interpretable for metallic/insulating character at a glance.

Validation & Pitfalls

  • A phase diagram figure showing entries without their energy-above-hull value invites a reader to assume all shown phases are equally accessible/stable — always provide hull-distance context (via labels, color-coding, or a stated cutoff for what's shown), consistent with materials-structure-analysis/references/phase-diagrams.md's point that a small positive hull distance still may (or may not) be synthesizable.
  • Engineering vs. true stress-strain must be labeled explicitly on any plotted curve — the same pitfall as in mechanical-testing, but specifically a figure-labeling failure mode: an unlabeled axis lets a reader assume whichever convention they're used to, which can be wrong.
  • Color scale and axis range choices can visually exaggerate or hide effect size, the same general principle as this repo's sibling neuroscience figures skill — choose ranges justified by the data's actual distribution, and state them, rather than defaulting to auto-scaled ranges that may mislead about magnitude.
  • DFT-computed and experimentally measured quantities (e.g. a computed vs. measured band gap) must be visually distinguishable when shown together — using identical visual treatment for a computed prediction and a measured value in the same figure invites equal-confidence interpretation, which the systematic DFT band-gap underestimate (see computational-materials/references/dft.md) doesn't support; use distinct markers/line styles and label which is which.
  • Simulated/projected results (a candidate material's predicted property, an extrapolated trend) must be visually distinguished from measured results — matching this repo's sibling neuroscience figures skill's principle exactly: use a distinct visual treatment (dashed lines, hatching, explicit labeling) for anything that isn't a direct measurement.

示例提示词

帮我按发表规范画一张带极图的相图和能带DOS组合图。

所属插件

材料表征、性能与材料信息学面向实验与数据驱动的材料研究:XRD/电镜/谱学/热分析/力学测试等表征,合成与加工,电化学与半导体表征,相场、CALPHAD 与连续介质力学,材料数据库、特征工程、贝叶斯优化与主动学习,选材与生命周期评价。

安装该插件会一并获得同领域的 42 个 Skill / MCP。

同插件的其他科研 Skill

常见问题

材料科学图表绘制规范能做什么?

Covers publication-quality figure conventions for materials science plots: phase diagrams, pole figures, stress-strain curves. Use when the deliverable is a figure, not just an analysis result。

如何安装材料科学图表绘制规范?

在科搭 Codar 左侧点击「插件」,在「组件」页签搜索「材料科学图表绘制规范」并安装;在 Claude Code 或 Codex 中可以发送本页的安装提示词,或下载 Skill 包后解压到 ~/.claude/skills 或 ~/.codex/skills。

材料科学图表绘制规范需要哪些权限?

该科研 Skill没有声明额外权限;执行具体操作时仍由 Agent 按自身规则请求确认。

材料科学图表绘制规范经过官方审核吗?

是。SpeedSkills 收录的每个版本都会记录内容摘要与权限声明,当前版本已通过安全审核。