Python API

Nucleide exposes a typed pure-Python package at python/nucleide/ that re-exports a compiled PyO3 extension built from bindings/python/. The extension module is named nucleide._internal; users import from the nucleide domain submodules, which mirror the Rust workspace crates.

Module layout

import nucleide as nuc

The top level carries version(), nucleide.__version__, and the domain submodules below (so nuc.nuclei.Nuclide(...) works directly). Each submodule re-exports its symbols from nucleide._internal:

Submodule	Backing crate	Contents
`nucleide.nuclei`	`nuclei`	Nuclide/particle identifiers, nuclear data, reaction names
`nucleide.material`	`material`	Compositions, compendium, activity, XML export
`nucleide.mcnp`	`mcnp-io`	MCNP file readers and writers
`nucleide.serpent`	`serpent-io`	Serpent output readers
`nucleide.fluka`	`fluka-io`	FLUKA USRBIN readers
`nucleide.vr`	`vr-tools`	MAGIC weight windows, source sampling
`nucleide.enrichment`	`enrichment`	Enrichment cascades
`nucleide.depletion`	`depletion`	Depletion chains and CRAM solves
`nucleide.data`	— (pure Python)	Release-pinned data-file downloads

`nucleide.nuclei`

Nuclide(name) — canonical nuclide identifier with cross-code name conversions.
from_zaid(zaid) → Nuclide from an MCNP ZAID integer.
Particle(spec) — particle species with MCNP/FLUKA/Geant4 translations.
atomic_mass(key) — by nucid integer or name.
natural_abundance(key)
half_life(key)
decay_constant(key)
q_value_capture(key)
q_value_alpha(key)
rxname_id(name) → reaction id
rxname_name(id) → name
rxname_mt(id) → ENDF MT number

`nucleide.material`

from_formula(formula) → composition dict
MaterialsCompendium.load(path) → MaterialsCompendium
activity(comp) → activity dict
to_xml(comp, name, density, units="g/cm3") → materials XML string

`nucleide.mcnp`

read_xsdir(path) → Xsdir
read_meshtal(path) → Meshtal
read_wwinp(path) → Wwinp
read_mctal(path) → Mctal
read_ssw(path) → SurfSrc
read_ptrac(path) → PtracFile
write_ssw(ssw, path, tracks=None)
read_inp(path) → list of material dicts
mesh_to_geom(...) → MCNP geometry deck string

`nucleide.serpent`

read_serpent(path, kind) → structured dict

`nucleide.fluka`

read_usrbin(path) → list of UsrbinTally

`nucleide.vr`

magic(tally, per_group=False, tolerance=0.5) → MagicOutput
AliasTable(pdf) — Walker alias table for discrete sampling
MeshSourceSampler(tally, mode, user_pdf=None) — mesh source sampler

`nucleide.enrichment`

Cascade.default_uranium() → Cascade
c.solve(tolerance=None, max_iterations=None)
c.solve_multicomponent(tolerance=None, max_iterations=None) — M*-optimizing multicomponent solve

`nucleide.depletion`

read_chain(path) → Chain
build_depletion_system(chain, rates) → DepletionSystem
system.solve(n0, dt, order=48) → result dict
system.solve_vec(n0, dt, order=48) → result list in chain nuclide order
deplete(chain, n0, dt, rates=None, order=48) → result dict

`nucleide.data`

Pure-Python helpers (no backing crate) for downloading repo data files that the wheel does not bundle, pinned to the installed release:

fetch(path, ref=None, dest=".") → local path str of a repo-relative file (e.g. "fixtures/depletion/chain_simple.xml")
fetch_compendium(ref=None, dest=".") → local path str of the Materials Compendium JSON
default_ref() → the installed version’s tag (e.g. "v0.1.0"); pass ref="main" or a commit SHA to override

Version

version() → workspace version string
nucleide.__version__ — same value

For full type signatures, see python/nucleide/_internal.pyi.