Keyboard shortcuts

Press ← or → to navigate between chapters

Press S or / to search in the book

Press ? to show this help

Press Esc to hide this help

Versions and releases

Software, nuclear data, and desktop packages have separate release histories. Record all the versions relevant to your calculation.

ComponentVersion covered hereRelease
CLI, Python, and Rust workspace1.7.2v1.7.2
Embedded nuclear-data catalog1.2.0data-v1.2.0
Desktop preview0.6.0-preview.1, using solver 1.7.0desktop preview

Use actinv --version for the executable. actinv data manifest prints its embedded catalog. Source-built desktop candidates may use a newer solver than the published preview; inspect their package/build record.

New in 1.7.2

Release 1.7.2 makes actinv data fetch work for first-time installs again. The IAEA site now refuses automated downloads of the two decay files, and fetch now takes ACTINV’s byte-identical copy when that happens. The OpenMC R2S adapter also ships with the Python package (from actinv.openmc import ActinvR2SManager), with an opt-in uncertainty band on every photon tally; see Use ACTINV with OpenMC.

It also adds mixture self-shielding (self_shielding.mixture) and faster rate pruning, and fixes the potential-scattering estimate of dilution: "composition" self-shielding. The 1.7.2 release notes have the detail.

New in 1.7.1

Release 1.7.1 fixes resonance capture in libraries and self-shielding tables you build from R-matrix-limited evaluations. These include ENDF/B-VIII.1 and FENDL-3.2c tungsten, copper, Fe-54 and Fe-57; from 1.5.0 to 1.7.0 their thermal capture was off by factors up to 9. If you built such a library or table, rebuild it. The TENDL-2025 data that ACTINV downloads is unaffected.

It also fixes the study.robustness cross-section sampler and adds:

  • national clearance tables (actinv clearance --jurisdiction de|uk|ca);
  • a shutdown-dose source covariance export (actinv export-r2s-cov);
  • build-shielding --profile and --workers, and a profile for geometry self-shielding.

The 1.7.1 release notes list what to re-check.

New in 1.7.0

Release 1.7.0 adds impurity screening for fusion materials (actinv impurities), a damage table in the data catalog (1.2.0) so that dpa, exact NRT dpa, arc-dpa and He/dpa need no table to build, FISPACT-II parity additions (estimated photon spectra, fission-yield and uncertainty keywords in actinv import-fispact, and actinv build-damage --method kinematic), actinv new --material with a cited material library and standard spectra, and actinv report --format html. The desktop 0.6.0-preview.1 reaches most of the command line: an Export page, and pages for impurity screens, certification, optimization, comparison, waste questions, transport meshes and measurements; the browser sets these up and opens two results solved ahead of time. The pathway closure no longer counts noise-level populations, and actinv decide computes the measurement-design blocks it used to leave null. See What’s new in 1.7, the 1.7.0 release notes and the desktop 0.6.0-preview.1 notes.

New in 1.6.1

Release 1.6.1 fixes one defect of 1.6.0. A trace-mode run with pathway output, which the default outputs include, could stop with an index out of bounds error. This happened when a reaction of a bulk nuclide led to a product with no decay data. Runs that completed in 1.6.0 give the same results in 1.6.1.

New in 1.6.0

Release 1.6.0 adds waste classification under nuclear-data uncertainty (actinv waste uncertainty), impurity budgets under uncertainty (actinv waste budget uncertainty), classes for packages of discrete items with encapsulation (actinv waste package), the waste commands in the Python API (actinv.waste, actinv.twin) and a Waste class page in the desktop. Rate pruning now drops a nuclide only when a proven bound on its population is below the threshold, and the ledger’s rate_pruning block reports bounds on the change in total activity and heat; see What’s new in 1.6.

New in 1.5.0

Release 1.5.0 adds FISPACT-II deck import (actinv import-fispact), result tables (actinv report) and library queries (actinv inspect), a step summary and depletion table, photon dose at a distance with a separate bremsstrahlung block, production pathways and routes with closure, bundled regulatory tables (49 CFR 173.435/173.436 and Euratom 2013/59 Annex VII), geometry self-shielding, sampling statistics for robustness studies, wall-loading and power-driven steps, and spontaneous-fission products. The desktop 0.4.0-preview.1 adds the Outputs, Tables, Photons & dose and Pathways pages and deck import. Some defaults changed; see Changes that may affect earlier results. The data catalog stays at 1.1.0. What’s new in 1.5 is a tour of the additions; see also the 1.5.0 release notes and the desktop 0.4.0-preview.1 notes.

New in 1.4.0

Release 1.4.0 adds impurity-budget workflows (actinv budget and the Python budget helper), the OpenMC R2S adapter, hydrogen/helium isotope production through options.gas, transport-tally statistical error propagation through the flux uncertainty channel, lean mesh output through dotted steps.<field> entries in cell_result_fields, and photonuclear (incident-gamma) activation through build-library --projectile gamma. Gamma activation libraries are not published through actinv data fetch; build one from the TENDL g sublibrary. See the 1.4.0 release notes.

Source builds between releases can carry a workspace version that has not moved yet, so record the commit as well as actinv --version.

Upgrade software

python -m pip install --upgrade actinv

For an exact Rust CLI version:

cargo install --locked --force actinv-cli --version 1.7.2

A software upgrade does not silently replace installed data. Revisit optional settings and coverage when a release changes their behavior, and retain the executable or source version for older results.

Changes that may affect earlier results

The 1.7.2 release notes change no default result. Rerun calculations that used self_shielding with dilution: "composition" on materials whose main scatterers are light nuclides missing from the table (concrete, water); from 1.1.0 to 1.7.1 they were over-shielded.

The 1.7.1 release notes change no default result. Rebuild any library or self-shielding table you built from R-matrix-limited evaluations (ENDF/B-VIII.1 and FENDL-3.2c tungsten, copper, Fe-54, Fe-57 and others) with 1.5.0 to 1.7.0, and rerun what used it. study.robustness bands from cross_section_mf33 can also differ from 1.7.0.

The 1.7.0 release notes list the two default-result changes: pathway_closure counts only populations above 1e-12 of the step’s largest, so the quick start reports 3.9e-15 instead of 4.117, and actinv decide computes the design.top_reactions blocks it reported as ill_conditioned. Results that may change has the detail.

The 1.6.0 release notes describe the one default-result change: the new rate-pruning rule keeps or drops a few different noise-level nuclides, and the ledger’s rate_pruning fields are renamed and extended. Results that may change gives the measured effect.

The 1.5.0 release notes list the default-result changes: trace mode with radioactive initial material (restart or cooling runs from an activated inventory should be rerun), trace-mode pathway attribution, the spontaneous-fission branch count, and corrections to opt-in photon, regulatory-table and self-shielding outputs. Results that may change summarizes them.

The 1.2.1 release notes identify self-shielding interpolation, composition validation, and decay-branch accounting corrections. Review them if your earlier calculations used those paths.

The 1.3.1 release notes describe that release’s capabilities and their qualification boundaries. The 1.4.0 release notes identify a trace-mode correction that moves results for multi-element materials; review them if your earlier calculations used trace mode. The changelog retains the full software history; data-release notes describe data changes.

Publishing a new executable does not repair or supersede an evaluation’s recorded failure. See Known data limitations before assuming a newer software version implies better nuclear data.