johanjoensson / impurityModel
95%

DEFAULT BRANCH: master
Repo Added
Files 103
Badge
Embed ▾
README BADGES
x

If you need to use a raster PNG badge, change the '.svg' to '.png' in the link

Markdown

Textile

RDoc

HTML

Rst

LAST BUILD ON BRANCH master
branch: master
CHANGE BRANCH
x

coverage: 95.376% (+0.007%) from 95.369%
43448266298

push

github

johanjoensson 
Change fallback valence/conduction split

Instead of looking at bath on-site it looks at the bath only density
matrix and fills the bath orbitals according to the occupations. Still
coarser than just passing in a split calculated from the fitting.

23 of 24 new or added lines in 2 files covered. (95.83%)

110 existing lines in 12 files now uncovered.

21718 of 22771 relevant lines covered (95.38%)

0.95 hits per line

Relevant lines Covered
22771 RELEVANT LINES 21718 COVERED LINES
0.95 HITS PER LINE
Source Files on master

Recent builds

Builds Branch Commit Type Ran Committer Via Coverage
43448266298 master Change fallback valence/conduction split Instead of looking at bath on-site it looks at the bath only density matrix and fills the bath orbitals according to the occupations. Still coarser than just passing in a split calculated from the fitting. push johanjoensson github
95.38
43308190163 master tests: freeze_replay restores the module settings it overrides; workload benchmarks skip without WORKLOAD_H5 Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com> push johanjoensson github
95.37
43226228527 master Solver works in the basis it is given: remove the adaptive rotation prepare_solver_basis, calc_selfenergy and the spectra driver no longer rotate the impurity into a symmetry-adapted basis and back. Choosing the basis is the caller's job (rspt2sp... push johanjoensson github
95.32
43039006152 master CI: apt retries and 20 s timeouts, so a flaky runner mirror fails fast or recovers Every job stalled ~15 min in 'Install system dependencies' on a mirror outage (Err ... Ign over azure/archive/security). Co-Authored-By: Claude Sonnet 5.5 <norepl... push johanjoensson github
95.27
40071905347 master chore(lint): make ruff, black and mypy clean across the package `make check` was red on master after merging nio-iter1-findings: 138 ruff findings (125 of them pre-existing under ruff 0.15/0.17), one file black would reformat and 40 mypy errors (... push johanjoensson github
94.72
38922753214 master Merge pull request #5 from johanjoensson/massive_core_refactor Core engine redesign: one sparse type, one Lanczos step, 52 -> 35 knobs (draft) push web-flow github
94.49
38130571996 master Merge pull request #3 from johanjoensson/gf-queue-scheduler Gf queue scheduler Use a MPI distributed task queue for calculating the interacting Greens function, decreasing the MPI ranks idle times, hopefully pushing the average load up. Before i... push web-flow github
94.05
37474447036 master fix(gf): bound the frozen-CSR recurrence's first block budget; free the stopped sparse state block_lanczos_array preallocates its coefficient (and PARTIAL W) buffers at max_iter, and the fallback passed ceil(N/p): ~32 N p bytes, GiBs at a 10M fro... push johanjoensson github
93.96
37441504823 master docs(gf): normalised SrMnO3 state weights; gf_min_weight leaves thermal_rho whole The eight excited SrMnO3 states weigh 5.7-6.8e-5 each normalised (5.0e-4 in total), not the unnormalised 1.4e-4 quoted in 0e92e7b8's message. Also document that the... push johanjoensson github
93.94
37374948539 master test(basis): the block-merge ratio bound is release-calibrated A cluster run reported 10.5x against the 8.0 bound on current code. A safe build of the same tree measures 6.6-7.2x here against 4.1x for release, so the rise is the build, not a key-... push johanjoensson github
93.93
See All Builds (117)