• Home
  • Features
  • Pricing
  • Docs
  • Announcements
  • Sign In

Qiskit / qiskit-fermions / 33761644645
98%

Build:
DEFAULT BRANCH: main
Ran 03 Sep 2026 01:36PM UTC
Jobs 1
Files 108
Run time 1min
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

03 Sep 2026 01:31PM UTC coverage: 98.184%. Remained the same
33761644645

push

github

web-flow
Fix the Trotterization and the plaquette stabilizers in the flow-set guides (#309)

* fix(docs): correct the Trotterization and the plaquette stabilizers in the flow-set guides

Addresses the three problems reported in #306.

**Multi-step Trotterization.** `FlowSetSynthesis` repeated its gate sequence
internally, but it receives one flow set at a time whenever the circuit has been
decomposed first. The repetition then stayed within a set instead of cycling
between them, and that achieves nothing at all: consecutive exponentials of the
same operator commute, so `A(t/2) A(t/2)` is just `A(t)` and the error sat at its
one-step value however many steps were asked for.

In 2D the pipeline always decomposes beforehand, so `reps` could never have
interleaved the sets, and it is removed. The check that pinned it compared the
flow-set synthesis against LieTrotter with both sides fed one set at a time, so
both nested identically and the comparison passed while testing nothing about
multi-step Trotterization.

In 1D the convergence comparison passes the whole Hamiltonian in a single gate,
where the repetition is correct, so rather than remove a working benchmark the
synthesis now refuses `reps > 1` when handed a single flow set.

**The plaquette stabilizers.** These were built as a product of the VC-encoded
transfer loop with the Jordan-Wigner image of the same loop, described as
"composing them cancels the physical content". That description was wrong, and it
obscured what the second factor does: it is the local parity of the plaquette's
four sites. The bare transfer loop does not commute with the encoded Hamiltonian
-- the bonds that spoil it are exactly those leaving the plaquette -- because the
Hamiltonian hops particles across the boundary and so conserves only the global
parity. The parity factor is what repairs that.

Naming it that way makes the construction collapse. The library decomposes
`T_jk = (i/2) V_j E_jk`, so `E_jk = -2i V_j T_jk`, and going aro... (continued)

13412 of 13660 relevant lines covered (98.18%)

482880.82 hits per line

Jobs
ID Job ID Ran Files Coverage
1 33761644645.1 03 Sep 2026 01:36PM UTC 108
98.18
GitHub Action Run
Source Files on build 33761644645
  • Tree
  • List 108
  • Changed 0
  • Source Changed 0
  • Coverage Changed 0
Coverage ∆ File Lines Relevant Covered Missed Hits/Line
  • Back to Repo
  • Github Actions Build #33761644645
  • bae161c5 on github
  • Prev Build on main (#33646103237)
  • Next Build on main (#33862333505)
STATUS · Troubleshooting · Open an Issue · Sales · Support · CAREERS · ENTERPRISE · START FREE TRIAL · SCHEDULE DEMO
ANNOUNCEMENTS · TWITTER · TOS & SLA · Supported CI Services · What's a CI service? · Automated Testing

© 2026 Coveralls, Inc