Skip to content

Releases: TRIQS/solid_dmft

Version 3.3.1

29 Aug 18:59
Compare
Choose a tag to compare

solid_dmft version 3.3.1 is a minor releases that fixes the following
problems and or brings the following improvements:

  • align ctseg interface with latest release of triqs/ctseg
  • fixes a wrong access of solver params
  • fix CSC routines
  • fix initial DC when performing a magnetic calculation
  • add missing default.toml to pypi packing
  • switch to Ubuntu 24.04 in CI and github actions

Version 3.3.0

27 May 14:59
Compare
Choose a tag to compare

solid_dmft 3.3.0 is a major release, compatible with TRIQS 3.3, updated to the latest app4triqs skeleton, and bringing major changes to the code:

  • the input parser is switched to a general toml parser, i.e. strings have to be passed in quotes, boolean parameters given without capitalization, and lists passed with brackets. Check below separate section for detailed changes.
  • the new input parser allows to define for each impurity problem a different solver if wanted, i.e. d-shell with cthyb and p-shell with Hartree-Fock. See new text NiO-cthyb
  • docker images are automatically build on each push for all major releases to ghcr.io
  • switch from old ctseg to new ctseg_j solver
  • allow CRM Dyson solver for both cthyb and ctseg to obtain Sigma_imp
    from G_tau: "crm_dyson_solver=true" and dlr_wmax and dlr_eps (see https://triqs.github.io/triqs/unstable/documentation/python_api/triqs.gf.dlr_crm_dyson_solver.html#module-triqs.gf.dlr_crm_dyson_solver for details)
  • add new DC schemes 'crpa_static', 'crpa_static_qp', 'crpa_dynamic'
  • use cRPA calculated Uijkl as interaction via 'crpa',
    'crpa_density_density', 'dyn_density_density', 'dyn_full' hint types
  • read interaction tensor from AIMBES h5
  • new experimental gw_embedding module. See gw_embedding/gw_flow.py for details allowing to run solid_dmft on top of AIMBES
  • allow to use Pade for AC in post-processing

We thank all contributors: Sophie Beck, Thomas Hahn, Alexander Hampel, Henri Menke, Maximilian Merkel, and Nils Wentzell

A detailed list of changes is provided in the ChangeLog.

Version 3.2.3

16 May 13:47
Compare
Choose a tag to compare

solid_dmft version 3.2.3 and 3.2.2 are minor releases that fixes bugs in the post-processing routines and brings small new improvements:

  • allow 0.0 mixing to perform stat sampling
  • allow mathematical expression to be passed for random_seed
  • fix broken FS plot in PCB: np.shape -> len
  • fix PCB bug aprx Sigma as diagonal if interpolation is used
  • draw coloarbar only once in kslices

We thank all contributors: Alexander Hampel, Henri Menke

Version 3.2.0

01 Sep 11:22
Compare
Choose a tag to compare

solid_dmft version 3.2.0 is a release that

  • adds Jenkins CI support via flatiron-jenkins
  • includes several fixes to match the latest triqs 3.2.x release
  • changes the Z estimate to a correct linear fit of the first two Matsubara frequencies
  • fixes for QE and Vasp CSC
  • add option to add a magnetic field in DMFT
  • add solid_dmft JOSS paper reference (doi.org/10.21105/joss.04623)
  • add simple Ntot interaction
  • allow Uprime!=U-2J in Kanamori
  • updates the tutorials
  • introduces input output documentation
  • add support for the TRIQS Hartree Solver
  • add RESPACK support

We thank all contributors: Sophie Beck, Alberto Carta, Alexander Hampel, Max Merkel, Harrison LaBollita, Nils Wentzell

A detailed list of changes is provided in the ChangeLog.

Version 3.1.5

20 Jun 16:25
Compare
Choose a tag to compare

solid_dmft version 3.1.5 is a patch-release that improves / fixes the following issues:

  • fix to charge-self-consistency with Vasp and QE
  • feat add loc_n_min / max arg for cthyb
  • fix simple Z estimate as linear fit
  • adapt docker images for ubuntu 22.04

Contributors: Sophie Beck, Alberto Carta, Alexander Hampel, Max Merkel

Version 3.1.4

23 Feb 20:32
Compare
Choose a tag to compare

solid_dmft version 3.1.4 is a patch-release that improves / fixes the following issues:

  • fix and improve rootfinder in PCB for quasiparticle dispersion
  • fix pypi package version.py module

Contributors: Sophie Beck, Alberto Carta, Alexander Hampel, Max Merkel:

Version 3.1.3

18 Jan 22:54
Compare
Choose a tag to compare

solid_dmft version 3.1.3 is a patch-release that improves / fixes the following issues:

  • fix delta interface of cthyb for multiple sites with different block structures
  • correctly use tail fitted Sigma from cthyb not via double dyson equation
  • magnetic param not available in CSC crash PM calc
  • improve PCB script from unstable branch
  • convert warmup cycles to int automatically
  • fix function calls in gap finder
  • fix delta_interface with SOC and store solver options
  • fix: update svo example for PCB test from unstable

Contributors: Sophie Beck, Alberto Carta, Alexander Hampel, Max Merkel:

Version 3.1.2

22 Sep 13:40
Compare
Choose a tag to compare

solid_dmft version 3.1.2 is a patch-release that improves / fixes the following issues:

  • fix deg shells in magnetic calculations
  • fix bug in max_G_diff function ignoring norm_temp
  • fix load Sigma with new gf_struct in triqs 3.1.x
  • Made the way mesh is stored in maxent_gf_latt consistent with maxent_gf_imp
  • adapt vasp patch files for ver6.3.2
  • update README.md for Joss publication
  • print optimal alpha in other maxent scripts
  • update postprocessing routines for plotting spectral functions
  • add new test for pcb get_dmft_bands function
  • DOC: extend install instructions & improve readme for #21 #22
  • DOC: update support & contribute section, bump ver to 3.1.1
  • add proj_on_orb functionality to Akw
  • Added observables documentation for DMFT output
  • Added input documentation
  • Added ETH logo to website, small fixes to documentation
  • rename examples to debbuging_examples
  • pip package build files

Contributors: Sophie Beck, Alberto Carta, Alexander Hampel, Max Merkel

Version 3.1.1

07 Jun 00:57
Compare
Choose a tag to compare

solid_dmft version 3.1.1 is a patch-release that improves / fixes the following issues:

  • delete obsolete make_spaghetti.py
  • SOC self energies can be continued in maxent
  • run hubbardI solver on all nodes due to slow bcast performance of atomdiag object
  • fix DFT energy read when running CSC QE
  • updated documentation, small fixes to tutorials
  • exposed params of maxent_gf_imp
  • fix the way dft_mu is loaded in PCB
  • fix executable in SVO tutorial
  • fix shift in sigma continuator to remove dft_mu
  • fix chemical potential in plot Akw and minor fixes
  • correct plotlabels in postprocessing
  • tiny modification of printing H_loc in postprocessing

Contributors: Sophie Beck, Alberto Carta, Max Merkel

triqs 3.1.0 compatibility release

22 Apr 19:19
Compare
Choose a tag to compare
fix triqs ver to 3.1.x