Qiskit / qiskit-fermions / 34129339509
98%

DEFAULT BRANCH: main
Ran
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
coverage: 98.299% (+0.009%) from 98.29%
34129339509

push

github

web-flow 
docs: explain the relationship between qiskit-fermions and ffsim (#327)

The feedback that started this stack was that the difference between the two
packages is not clear enough, and the docs were a cause as much as the code
was. The best statement of the relationship already existed, but it sat in a
guide titled "The ffsim simulation backend" filed under "Workflow validation",
which is nowhere a reader wondering about the difference would look. The
landing page introduced ffsim with a single bullet about "integrating with
other tools of the ecosystem", saying nothing about the division of labor.

So the explanation now appears at all three places a reader can start:

- The landing page and README gain a "Relationship to ffsim" section stating
  the split: ffsim owns the ansatz operators and the simulation, this package
  owns the fermionic circuit and its transpilation under any encoding.
- The ffsim guide is retitled "How this package relates to ffsim" (anchor
  renamed to match, along with its inbound references) and opens with a table
  of who owns what before explaining the simulation protocols.
- The LUCJ guide is recast from "Build an LUCJ ansatz" to "Transpile an LUCJ
  ansatz", which is what it is for now that ffsim builds the operator. It moves
  out of "Circuit examples > Chemistry" into "Transpiling fermionic circuits",
  with its anchor renamed from lucj_getting_started to lucj_transpilation (five
  inbound references updated).

That recast is a restructure, not a retitle. Most of the old guide duplicated
ffsim's own qiskit-lucj guide step for step, differing only in substituting
UCJ(ucj_op) for UCJOpSpinBalancedJW. The molecular Hamiltonian construction,
the energy evaluation and the compressed-double-factorization section are
therefore gone rather than migrated: the first two demonstrate ffsim's
protocols rather than anything here, and the third reduced to passing a
different operator to the same constructor. What is left is only what ... (continued)

11208 of 11402 relevant lines covered (98.3%)

578099.53 hits per line

Jobs
ID Job ID Ran Files Coverage
1 34129339509.1 108
98.3
GitHub Action Run
Source Files on build 34129339509