Fix VaporPhase mass-basis latent heat weighting

[andres9403]

Summary

• Fix VaporPhase.getEnthalpy(basis="mass") so the per-species J/kg latent heat term is weighted by mass fractions.
• Preserve the existing mole-basis weighting and the total_h=False per-species return path.
• Add a regression test with two species of very different molecular weights that verifies mass-basis/mole-basis consistency.

Closes #18

Tests

• pytest tests/test_phases_vapor_enthalpy.py
• pytest tests/ -m "not assimulo"
• pytest tests/ --collect-only -q
• git -c core.whitespace=blank-at-eol,blank-at-eof,space-before-tab,cr-at-eol diff --cached --check

Assimulo-marked integration tests were not run locally because Assimulo is not installed in this environment; the CI workflow runs that marker group separately.

Branch Hygiene

• Base branch: master
• Source branch point: origin/master at 1c3a046fceb5c4da8b5fdb6e8b254f657d6e3cc7
• Upstream status: origin/master includes upstream/master at 400e1e13ad4ecd3ba202d5fae8c70ec73bdf9c3e
• Stacked status: not stacked; no prerequisite PRs
• Note: an old remote branch named fix-issue18-latent-heat-weighting exists at an ancestor of current origin/master and has no PR; this PR uses a fresh focused branch.

[andres9403]

Maintainer review — PR #100 (VaporPhase mass-basis latent heat weighting)

Reviewed at head 56c6835. The change correctly resolves issue #18: in VaporPhase.getEnthalpy, the per-species J/kg latent-heat term is now weighted by mass fractions when basis='mass' and by mole fractions otherwise, instead of using mole_frac unconditionally.

What I verified

• Fix matches the diagnosis. PharmaPy/Phases.py:732-733 selects frac = mass_frac if basis == 'mass' else mole_frac, exactly the fix recommended in the issue audit and the independent Codex verification. mass_frac/mole_frac are both resolved earlier (lines 687-693), so the mass branch is always defined.
• Mole-basis path unchanged. The else branch still uses mole_frac; the total_h=False per-species path (hVap = hSens + deltaVap) is untouched. All existing callers in Evaporators.py/Extractors.py pass basis='mole' explicitly and are unaffected.
• Regression test is meaningful. tests/test_phases_vapor_enthalpy.py uses two species of disparate MW (18 vs 100), equimolar, and independently recomputes the expected mass- and mole-basis mixture enthalpies plus the basis-invariance identity h_mass == h_mole * 1000 / mw_av, and checks the per-species path for both bases. I confirmed it is genuinely red on the pre-fix code shape (1.51e6 vs expected 9.46e5 J/kg) and green after the fix — it locks in the behavior rather than passing vacuously.
• Docstring cleanup fixes the he default is True typo and documents the basis parameter, consistent with the signature default ('mass').

Tests run locally (Python 3.10, numpy 1.26.4)

• pytest tests/test_phases_vapor_enthalpy.py → 1 passed
• pytest tests/ -m "not assimulo" → 28 passed, 6 deselected
• Temporary revert of the one-line fix → target test fails as expected, then restored
• Working tree clean; no stray artifacts. Assimulo-marked tests not run (Assimulo not installed locally; CI runs that marker group).

Findings

Nonblocking — pre-existing multi-temperature mass-basis shape mismatch (out of scope).
Not introduced by this PR, but adjacent to the changed line: in getHeatVaporization the basis='mass' branch reindexes deltahvap to only the subcritical columns (deltahvap[:, idx] / self.mw[idx]), reducing its second dimension from num_comp to len(idx). When any species is supercritical, the subsequent np.dot(deltaVap, frac) in getEnthalpy would then mismatch the full-length frac vector. The tests here keep all species subcritical (350 K), so this path isn't exercised. Worth a follow-up issue; no change requested in this PR.

Merge-readiness

• No blocking findings. Fix is correct, minimal, and covered by a regression test that fails without it.
• mergeStateStatus: CLEAN.
• This review is a COMMENT because I am the PR author — GitHub does not accept an APPROVE from the author account. If branch protection requires a formal approval, another eligible maintainer will need to provide it; live state does not currently report a REVIEW_REQUIRED gate.

[andres9403]

Addressed the review pass for PR #100.

Commits pushed:

• None; the branch was already up to date at 56c68351166867a09d3cdae8eb11cf029303c61f.

Main changes made:

• No PR code changes were needed. There are no unresolved inline review threads or discussion comments.
• The only review-body finding was explicitly nonblocking and out of scope for this PR: gh-review-pr:finding=multitemp-mass-getheatvap-shape.
• Recorded that pre-existing follow-up as High: [Phases] Multi-temperature VaporPhase mass-basis latent heat drops component columns #101: High: [Phases] Multi-temperature VaporPhase mass-basis latent heat drops component columns #101

Verification:

• Reproduced the follow-up behavior: getHeatVaporization(np.array([660.0, 680.0]), basis="mass") returns shape (2, 1) while the mole-basis path returns (2, 2), and getEnthalpy(..., basis="mass") raises ValueError: shapes (2,1) and (2,) not aligned.
• pytest tests/test_phases_vapor_enthalpy.py -> 1 passed.

Comments intentionally not addressed in this PR:

• The multi-temperature mass-basis shape mismatch was not fixed here because the maintainer review marked it nonblocking, pre-existing, and out of scope for PR Fix VaporPhase mass-basis latent heat weighting #100. It is now tracked in High: [Phases] Multi-temperature VaporPhase mass-basis latent heat drops component columns #101.

Remaining gates:

• Current PR head remains 56c68351166867a09d3cdae8eb11cf029303c61f.
• No formal approval is present yet; the existing maintainer review is a COMMENT by the PR author account.

[andres9403]

Review-resolution verification — no further review round needed

Verified at head 56c6835 (read-only pass).

• Inline review threads: none.
• Review-body findings: one finding (multitemp-mass-getheatvap-shape), explicitly nonblocking and out of scope for this PR; correctly deferred to tracking issue High: [Phases] Multi-temperature VaporPhase mass-basis latent heat drops component columns #101 (OPEN). Not a declined blocking comment.
• Core PR fix confirmed present at head: Phases.py still selects frac = mass_frac if basis == "mass" else mole_frac; regression test tests/test_phases_vapor_enthalpy.py present. No later commit reverted the fix.
• CI: Core tests pass, Assimulo integration tests pass; mergeStateStatus: CLEAN.

Another review round is NOT justified — the only finding was nonblocking and deferred by the reviewer, and no behavior changed beyond the commented lines. No further agent review needed.

Remaining gate (governance, not a review comment): reviewDecision is empty — the only review is a COMMENT from the PR author account, so a formal APPROVE from a second eligible maintainer may be required if branch protection demands one.