Die Oberfläche ist lokalisiert; Skill-Beschreibungen und -Dokumente werden per KI automatisch in Ihre Sprache übersetzt. Code, Befehle, Pfade und Fachbegriffe bleiben im Original, maßgeblich ist das Dokument des Autors.
Skills entdecken
Produktivität · SKILL.md

structures-local-environments-and-transformations

structures-local-environments-and-transformations

Construct and analyze pymatgen structures, molecules, local environments, graphs, dimensionality, prototypes, magnetism, molecule groups, and safe transformation-adjacent workflows.

Independent GitHub snapshot · Automated classification · Original documentation governs

Skill-Dokumentation

GitHub

name: structures-local-environments-and-transformations description: "Construct and analyze pymatgen structures, molecules, local environments, graphs, dimensionality, prototypes, magnetism, molecule groups, and safe transformation-adjacent workflows." disable-model-invocation: true metadata: disco-role: operating license: NOASSERTION

Structures, Local Environments, and Transformations

Use this sub-skill when the task is about Structure, Molecule, Lattice, local coordination, near neighbors, CrystalNN, VoronoiNN, MinimumDistanceNN, StructureGraph, dimensionality, structure matching, molecule matching, functional groups, Chemenv coordination environments, prototypes, or collinear magnetic moments.

Read First

  • For supported imports, inspected signatures, namespace-split notes, and API selection guidance, read references/api-reference.md.
  • For common in-memory recipes and diagnosis flows, read references/workflows.md.
  • For predictable failures and recovery steps, read references/troubleshooting.md.
  • To check a minimal runtime without source fixtures, run python scripts/structure_neighbor_smoke.py --help, then python scripts/structure_neighbor_smoke.py.

Routing Boundaries

  • Stay here for constructing/manipulating structures or molecules as prerequisites to local-environment, graph, dimensionality, prototype, magnetism, molecule-matching, functional-group, and Chemenv workflows.
  • Route computed entries, compatibility corrections, phase-diagram entries, batteries, and Borg/VASP assimilation to entries-thermodynamics-and-batteries.
  • Route slabs, surfaces, interfaces, Wulff shapes, Pourbaix, and interfacial reactivity to surfaces-interfaces-and-electrochemistry.
  • Route full pmg command syntax, persistent config, environment reporting, and POTCAR setup to cli-and-configuration.

Default Approach

  1. Build or load a valid Structure or Molecule; confirm coordinates, units, ordering, and oxidation states before analysis.
  2. Choose and name the neighbor/analysis model: start with CrystalNN, compare against MinimumDistanceNN or VoronoiNN when results are surprising.
  3. Preserve site indices and periodic image information when reporting neighbors or graph edges.
  4. Build a StructureGraph before calling graph or dimensionality functions.
  5. Explain optional dependencies and backend choices before using Chemenv, OpenBabel-backed functional groups, moyopy, or external aflow.

Important Compatibility Note

Current pymatgen installs and uses pymatgen-core for core objects and several compatibility-backed analysis modules. Use public imports from pymatgen.core and pymatgen.analysis.* as shown in references/api-reference.md; do not import Structure or Molecule from the root pymatgen package. If a class imported from pymatgen.analysis.local_env, pymatgen.analysis.graphs, or pymatgen.analysis.structure_matcher reports a pymatgen.core.* implementation module, treat that as expected compatibility behavior, not a bug.

Bundled Runtime Assets

  • references/api-reference.md covers core objects, neighbor strategies, graphs, structure matching, dimensionality, magnetism, prototypes, molecule workflows, Chemenv, and package-split caveats.
  • references/workflows.md provides copy-adaptable recipes for structures, molecules, neighbor comparison, graph dimensionality, Chemenv, magnetism, prototypes, matching, and serialization.
  • references/troubleshooting.md gives symptom-driven fixes for namespace split, oxidation states, neighbor disagreement, periodic images, malformed structures, optional extras, Chemenv/prototype backends, structure matching, and unit mistakes.
  • scripts/structure_neighbor_smoke.py builds tiny in-memory structures and prints deterministic local-environment and dimensionality diagnostics; it performs no network access, fixture reads, destructive writes, GUI opens, or long scans.
Gespeicherte Skills