rgpycrumbs.geom.fragments

Core fragment detection algorithms for molecular systems.

Provides geometric (scaled covalent radii) and bond-order (GFN2-xTB) based fragment detection, plus distance-based fragment merging.

Added in version 0.0.6.

Attributes

tbliteinterface
DEFAULT_BOND_MULTIPLIER
DEFAULT_BOND_ORDER_THRESHOLD

Classes

StrEnum	Backport for Python 3.10.
DetectionMethod	Available detection methodologies.

Functions

find_fragments_geometric(→ tuple[int, numpy.ndarray])	Detect molecular fragments using scaled covalent radii.
find_fragments_bond_order(→ tuple[int, numpy.ndarray, ...)	Analyze connectivity via the Wiberg Bond Order (WBO) matrix.
build_graph_and_find_components(→ tuple[int, ...)	Identify connected components using direct CSR sparse matrix construction.
merge_fragments_by_distance(→ tuple[int, numpy.ndarray])	Merges fragments with geometric centers closer than the specified distance.

Module Contents

class rgpycrumbs.geom.fragments.StrEnum

Bases: str, enum.Enum

Backport for Python 3.10.

rgpycrumbs.geom.fragments.tbliteinterface

class rgpycrumbs.geom.fragments.DetectionMethod

Bases: enum.StrEnum

Available detection methodologies.

Added in version 0.0.6.

GEOMETRIC = 'geometric'

BOND_ORDER = 'bond-order'

rgpycrumbs.geom.fragments.DEFAULT_BOND_MULTIPLIER = 1.2

rgpycrumbs.geom.fragments.DEFAULT_BOND_ORDER_THRESHOLD = 0.8

rgpycrumbs.geom.fragments.find_fragments_geometric(atoms: ase.atoms.Atoms, bond_multiplier: float, radius_type: str = 'natural') → tuple[int, numpy.ndarray]

Detect molecular fragments using scaled covalent radii.

Added in version 0.0.6.

rgpycrumbs.geom.fragments.find_fragments_bond_order(atoms: ase.atoms.Atoms, threshold: float, charge: int, multiplicity: int, method: str = 'GFN2-xTB') → tuple[int, numpy.ndarray, numpy.ndarray, numpy.ndarray]

Analyze connectivity via the Wiberg Bond Order (WBO) matrix. Calculate electronic structure using the specified xTB level.

Added in version 0.0.6.

rgpycrumbs.geom.fragments.build_graph_and_find_components(num_atoms: int, row_indices: numpy.ndarray | list[int], col_indices: numpy.ndarray | list[int]) → tuple[int, numpy.ndarray]

Identify connected components using direct CSR sparse matrix construction.

Added in version 0.0.6.

This function avoids Python list overhead by passing interaction indices directly to the SciPy sparse engine.

rgpycrumbs.geom.fragments.merge_fragments_by_distance(atoms: ase.atoms.Atoms, n_components: int, labels: numpy.ndarray, min_dist: float) → tuple[int, numpy.ndarray]

Merges fragments with geometric centers closer than the specified distance.

Added in version 0.0.6.