Releases: JuDFTteam/aiida-kkr
Releases · JuDFTteam/aiida-kkr
v2.3.1
Check the full changelog for important changes: v2.1.0...v2.3.1
v2.1.0
Important changes:
- add support for BdG calculations in impurity code
- add
imp_BdG
workflow - some bugfixes
- allow to deactivate SOC in impurity run with new solver
- update tests to aiida-core v2.3.0
- update dependencies
- improve LDA+U capabilities
v1.1.13: Merge pull request #117 from JuDFTteam/develop
Release 1.1.13
v1.1.12: Merge pull request #109 from JuDFTteam/develop
v1.1.11.post1: Remove aiida-testing dependency from testing extras
Otherwise upload tp pypi fails
Bugfixes
Merge branch 'develop', drop python 3.4 support, bumpversion -> v1.1.10
Dropped support for python 3.4 since the latest AiiDA-core requires graphviz==0.13 which is not available there
Release for aiida-core v1.0.0
v1.1.9 Merge branch 'develop', bumpversion -> v1.1.9
Fix pypi publishing
Fixed publishing step to pypi (poetry
forgot about the entrypoints).
Unification of workflow input nodes
Important changes:
- consistent naming conventions for workflow input nodes:
- computer options:
options
(respective entries are always:'queue_name', 'resources', 'max_wallclock_seconds', 'custom_scheduler_commands' , 'use_mpi'
) - names for JuKKR code input nodes :
kkr
(KKRhost),voronoi
,kkrimp
- if there is only one RemoteData input node for the workflow, the name should be
remote_data
, for more than one:remote_data_{...}
- general workflow parameter input nodes should always be named
wf_parameters
- the impurity information node for impurity embedding is always named
impurity_info
- computer options:
- consistent naming conventions for workflow output nodes:
- general information about the workflow run should be stored in a
workflow_info
ParameterData node
- general information about the workflow run should be stored in a
- fix auto tests