界面已本地化;技能介绍与说明文档按所选语言由 AI 自动翻译,代码、命令、路径和专业术语保留原文,以原作者文档为准。
探索技能
数据与研究 · SKILL.md

ovito-evidence-postprocessing

ovito-evidence-postprocessing

【数据与研究】Evidence-first OVITO atomistic postprocessing for LAMMPS trajectories and other particle data, including pipeline construction, modifiers, coloring, clipping, camera setup, images, animations, tables, and state files. Use when Codex must inspect or present molecular-dynamics results while distinguishing OVITO Basic GUI availability from ovitos/extern…

独立 GitHub 快照 · 自动分类 · 以原作者文档为准

技能说明

GitHub

name: ovito-evidence-postprocessing description: Evidence-first OVITO atomistic postprocessing for LAMMPS trajectories and other particle data, including pipeline construction, modifiers, coloring, clipping, camera setup, images, animations, tables, and state files. Use when Codex must inspect or present molecular-dynamics results while distinguishing OVITO Basic GUI availability from ovitos/external Python automation and separating rendering evidence from upstream solver evidence.

OVITO Evidence Postprocessing

OVITO explains and renders atomistic data; it does not prove that the upstream simulation was physically correct.

Start

ai-cae-toolbox context-scope --solver ovito
ai-cae-toolbox bridge-plan --solver ovito --objective "<objective>"
ai-cae-toolbox create-run --solver ovito --case <case-name> --objective "<objective>"
ai-cae-toolbox write-smoke-template --solver ovito --run-dir <run-dir>

Use ovito_check_installation before scripting. OVITO Basic may include ovito.exe but not ovitos; configure a compatible external Python with the OVITO package when batch scripting is required.

Build the Pipeline

Record input trajectory, frame range, particle properties, modifiers and their parameters, selection/deletion rules, color mapping, clipping plane, camera, render size, background, and export interval.

For strain or defect visualizations, record the reference configuration, neighbor cutoff, crystal structure, and modifier assumptions. Read references/render-evidence.md.

Finish

Export the pipeline script, PNG/MP4, numeric table when available, and a short caption defining every color and scalar range. State whether the output came from OVITO, a compatible Python implementation, or another plotting library. Never label a substitute plot as native OVITO output.

我的收藏