RMLE 1.0.0 — Stable algorithmic implementation for "Regularized Unfolding of gamma-ray Spectra for Nuclear Physics Applications"

Overview

This release (rmle-1.0.0) contains the finalized implementation of the Regularized Maximum-Likelihood Estimator (RMLE) algorithm described in the accompanying paper:

"Regularized Unfolding of gamma-ray Spectra for Nuclear Physics Applications"
by Erlend Lima, et al.

The RMLE algorithm provides a statistically consistent and numerically stable method for unfolding γ-ray spectra within the Oslo Method framework.

---

Contents

The RMLE method is implemented in the src/ompy/unfolding/rmle submodule.

Other parts of the OMpy package are still under development and are not part of this frozen algorithm release.

OMpy v2.1.0 (development preview)

This archive (response.tar.gz) is provided for internal testing and validation of data download workflows. Do not rely on its contents for production use.

Final public data releases will be announced separately.

Ompy v1.1.0

Most important changes:

• Changed response function interpolation between Compton edge and he chosen max. energy. Before, there was a misunderstanding of the bin-by-bin interpolation in Guttormsen1996. It now used a fan-like interpolation, too. Most noticable for response functions with small fwhm (like 1/10 Magne recommends for unfolding).

OMpy v1.0.0

Several changes, most important:

theoretical framework:

• Corrected likelihood: Forgot non-constant K term when K=K(theta): c8c0046

parameters and choices:

• unfolding parameters: 0fcafe2
• unfolder and fg method use remove negatives only, no fill: 9edb485
• Randomize p0 for decomposition 77dec9d

implementation and convenience:

• different save/load for vectors e5f7e52 and parent commits
• Enable pickling of normalizer instances via dill: 896b352

OMpy v0.9.1

Changes:

• corrected version number
  (v 0.9.0 has still v.0.8.0 as the version number)

OMpy v0.9

Many changes to the API have occured since v.0.2. Here a (incomplete) summary of the main changes:

• Vector and Matrix are now in mid-bin calibration. Most or all other functions have been adopted.
• Many changes (bugfix & readability) to the ensemble, decomposition and normalization classes.
• Normalization of nld and gsf ensembles working
• Parallelization, even though it could be more efficient for multinest (see #94 )
• Renamed response functions submodule; run git submodule update --init --recursive after git pull to get the new files
• remember to run pip install -e . such that changes to the cython files will be recompiled
• Documentation now available via https://ompy.readthedocs.io
• Installation requirements are hopefully all specified; docker file is provided with integration at https://hub.docker.com/r/oslocyclotronlab/ompy and mybinder can be used to rund the examples.
• We have clean-up the history of the repo to downsize it.
  Here the warning message: NB: Read this (only) if you have cloned the repo before October 2019: We cleaned the repository from old comits clogging the repo (big data files that should never have been there). Unfortunetely, this has the sideeffect that the history had to be rewritten: Previous commits now have a different SHA1 (git version keys). If you need anything from the previous repo, see ompy_Archive_Sept2019. This will unfortunately also destroy references in issues. The simplest way to get the new repo is to rerun the installation instructions below.

OMpy v0.2-beta

This is the first public beta version of the OMpy library, the Oslo Method in Python.

NB: Read this (only) if you have cloned the repo before October 2019 (which affects this release, v0.2-beta):
We cleaned the repository from old comits clogging the repo (big data files that should never have been there). Unfortunetely, this has the sideeffect that the history had to be rewritten: Previous commits now have a different SHA1 (git version keys). If you need anything from the previous repo, see ompy_Archive_Sept2019. This will unfortunately also destroy references in issues. The simplest way to get the new repo is to rerun the installation instructions below.

In essence: This tag does not work any longer; you have to download the version from https://zenodo.org/record/2654604