(changelog)= # Changelog ## Version 3.3.0 maxent version 3.3.0 is a compatibility release for TRIQS version 3.3.0 We thank all contributors: Thomas Hahn, Alexander Hampel, Henri Menke, Dylan Simon, Nils Wentzell Find below an itemized list of changes in this release. ### General * Try to speed up tests on macOS * Create font cache ahead of time * Fix documentation build * Removing no-triqs support ### doc * Minor corrections to sideb.html and conf.py.in * Remove redundant doc/sphinxext/numpydoc ### jenkins * Rename project to maxent * Disable sanitize build ## Version 3.2.0 Merge app4triqs skeleton into triqs_maxent ## Version 1.2.0 Version 1.2.0 is a compatibility release for TRIQS version 3.2.0 We thank all contributors: Alexander Hampel, Dylan Simon, Nils Wentzell Find below an itemized list of changes in this release. ### General * Fix issue in DataDefaultModel constructor for omegas of different length * Fix import issue for Python 3.10: Sequence should be imported from collections.abc * Synchronize Jenkinsfile and Dockerfile with app4triqs@a27e231c * Bump tolerance in SVD check of tau_kernel test from 1e-14 to 1e-13 * Porting to triqs 3.2 ### cmake * Bump maxent version to 1.2 and use triqs version 3.2 ### doc * Fix incorrect sign in expression for imaginary time Green function and kernel ### jenkins * Use gcc 13 on osx * remove bouncing email address * add CMAKE_ARGS in Dockerfile; fix node label for notriqs * use triqs unstable branch? ## Version 1.1.1 Fix an import issue for Python 3.10: Sequence should be imported from collections.abc ## Version 1.1.0 Version 1.1.0 is a compatibility release for TRIQS version 3.1.0. ## Version 1.0.0 Version 1.0.0 is a compatibility release for TRIQS version 3.0.0 that introduces compatibility with Python 3 (Python 2 no longer supported) ## Version 0.9.1 * :py:class:`.IOmegaKernel` for continuing :math:`G(i\omega)` (an object corresponding to :py:class:`.TauMaxEnt` is still missing) * :py:class:`.ComplexChi2` and :py:class:`.ComplexPlusMinusEntropy` for continuing complex :math:`A(\omega)` (Note: with the tau kernel, for single matrix elements and the elementwise formalism, it is usually not necessary to continue complex spectral functions in one go. Therefore, this is only interesting for using the :py:class:`.IOmegaKernel` or for the full matrix formalism which will be provided.) * Introduction of the :py:class:`.VerbosityFlags` and a new verbosity management system (as `suggested `_ by Marcel Klett) * Make package compatible with TRIQS release 2.1.0. The package might work with TRIQS 2.0, but we do not provide support for it. This version can be obtained using ``git checkout 0.9.1``. It has been checked to work with the following TRIQS releases: - TRIQS 1.4.2 release, hash `e7217b45e1e96fcf2820fd655346a057f22233d4 `_ - TRIQS 2.1.0 release, hash `e96f1d26217af753191482398dc568ac43aa920a `_ but it probably also works with other commits. ## Version 0.9.0 This is the initial version that was published. It includes (among other features) * analytic continuation with a :math:`G(\tau)` kernel, * the Bryan and MaxEnt cost function, * entropies for diagonal and off-diagonal elements, * elementwise and poor man's matrix MaxEnt, * several analyzers for determining :math:`\alpha`, * the sigma continuator, * tools for getting :math:`G(\omega)`, * plotting h5 files, and * preblur. This version can be obtained using ``git checkout 0.9``. It has been checked to work with the following TRIQS hashes: - TRIQS 1.4.x, hash `816aff2882e581b7fe0ae071842b53fc27d31346 `_ - TRIQS 2.0.x, hash `dc2cbc572478bcd5c35ba4777f6bbdeacd3f2262 `_ but it probably also works with other commits.