算法艺术创作
algorithmic-art
通过代码和可调参数创作生成艺术与交互式视觉效果。
visualization-utilities【设计与创意】Guides agents using datamol visualization, molecule image rendering, lasso highlighting, dataframe display, parallel utilities, filesystem helpers, RDKit log controls, and environment diagnostics.
独立 GitHub 快照 · 自动分类 · 以原作者文档为准
Use this sub-skill when the task asks to render molecules, save SVG/PNG depictions, highlight substructures, display molecule dataframes, visualize existing conformers, run small datamol utility jobs, inspect fsspec paths, silence RDKit logs, or diagnose datamol/RDKit runtime behavior.
../molecule-io-prep/ first.../structure-generation/ first.../fingerprints-similarity/ first.dm.viz.to_image() or top-level dm.to_image() from SMILES or RDKit molecules.dm.viz.lasso_highlight_image() and dm.viz.match_substructure().outfile paths handled through fsspec.dm.viz.conformers() when optional notebook widgets are installed and conformers already exist.dm.parallelized(), dm.parallelized_with_batches(), dm.JobRunner, dm.utils.fs, dm.utils.perf.watch_duration, dm.utils.decorators.disable_on_os, and RDKit log/version helpers.references/api-reference.md for signatures, parameters, return types, utility helper notes, and ownership boundaries.references/workflows.md for SVG/PNG rendering, lasso highlights, dataframe rendering, parallel jobs, fsspec checks, RDKit log control, and diagnostics recipes.references/troubleshooting.md for headless rendering, invalid highlight indices, SVG/PNG/Pillow issues, notebook assumptions, fsspec URI errors, and joblib pickling failures.scripts/visualization_utility_smoke.py --help for a safe executable smoke test, or run it with --output-dir <dir> to render a tiny SVG and exercise utility helpers.../molecule-io-prep/ if the task starts from mixed files, tables, or potentially invalid SMILES.use_svg=True for deterministic text artifacts and use_svg=False only when the caller explicitly needs PNG/Pillow output.lasso_highlight_image() for publication-style outline highlights and to_image(highlight_atom=..., highlight_bond=...) or match_substructure() for simple RDKit grid highlights.scheduler="threads" or n_jobs=0/1 for closures and notebook workflows.dm.without_rdkit_log() and restore global logs with dm.enable_rdkit_log() after temporary global muting.