refactor: vectorize explicit stencils, extract shared latent heat in 1D solvers

[formeo]

Related to #45

What changed

New file: _latent_heat.py
Extracts the latent heat phase-transition logic that was duplicated in all
4 solvers into a single apply_latent_heat() function.

Explicit solvers (explicit_general.py, explicit_k.py)

• Replaced copy.deepcopy + per-element Python loops with numpy vectorized
  FDM stencil computation
• Interior points computed in one array operation instead of a for-loop
• Boundary conditions applied to the numpy array directly

Implicit solvers (implicit_general.py, implicit_k.py)

• Replaced the inlined latent heat block with apply_latent_heat() call
• Minor cleanup: use n = obj.num_points local variable, remove unused
  copy import

What didn't change

• Function signatures and return format (y, lheat)
• FDM stencil formulas (same equations, just vectorized)
• Latent heat logic (exact same branching, moved to shared function)
• Implicit solvers still use np.linalg.solve with a tridiagonal matrix
  (no scipy.linalg.solve_banded — that could be a separate PR)

Testing

All 25 existing tests pass. Test runtime dropped from ~118s to ~72s on
the same machine (the effect will be larger on bigger grids since the
vectorization benefit scales with num_points).

Stats

4 files changed, 110 insertions(+), 213 deletions(-)
1 new file: heatrapy/dimension_1/solvers/_latent_heat.py (52 lines)

[codecov]

Codecov Report

❌ Patch coverage is 86.51685% with 12 lines in your changes missing coverage. Please review.
✅ Project coverage is 72.79%. Comparing base (8097499) to head (2e183e0).
⚠️ Report is 1 commits behind head on develop.

Files with missing lines	Patch %	Lines
heatrapy/dimension_1/solvers/_latent_heat.py	60.00%	10 Missing ⚠️
heatrapy/dimension_1/solvers/explicit_general.py	95.23%	1 Missing ⚠️
heatrapy/dimension_1/solvers/explicit_k.py	95.23%	1 Missing ⚠️

@@             Coverage Diff             @@
##           develop      #73      +/-   ##
===========================================
+ Coverage    70.66%   72.79%   +2.13%     
===========================================
  Files           21       22       +1     
  Lines         2335     2272      -63     
===========================================
+ Hits          1650     1654       +4     
+ Misses         685      618      -67     

Files with missing lines	Coverage Δ
heatrapy/dimension_1/solvers/implicit_general.py	96.15% <100.00%> (+34.92%)	⬆️
heatrapy/dimension_1/solvers/implicit_k.py	96.00% <100.00%> (+35.58%)	⬆️
heatrapy/dimension_1/solvers/explicit_general.py	96.00% <95.23%> (+37.30%)	⬆️
heatrapy/dimension_1/solvers/explicit_k.py	96.00% <95.23%> (+37.30%)	⬆️
heatrapy/dimension_1/solvers/_latent_heat.py	60.00% <60.00%> (ø)

... and 2 files with indirect coverage changes

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[djsilva99]

@formeo thanks a lot for the contribution — the implementation looks solid and the code quality is great overall 👍

I have a few small requests before merging:

• Could you update the PR title to "refactor: vectorize explicit stencils, extract shared latent heat in 1D solvers" to follow Conventional Commits? This helps ensure it’s properly reflected when generating new releases.
• Please remove the "Closes 45" from the PR description. The correct use of typing is part of that issue, so it shouldn’t be closed by this PR alone.
• It would be great to add a test covering a phase transition scenario. For example, you could create a simple integration test in test_single_1d.py using water as the material, with an initial temperature below 0ºC and a fixed temperature boundary condition above 0ºC. This would also increase the overall coverage that must be addressed.

Thanks again — really nice work!

[formeo]

Thanks Daniel! All three addressed:

1. ✅ PR title updated to follow Conventional Commits
2. ✅ Removed "Closes Refactor 1D solvers #45" from description
3. ✅ Added phase transition test using water (263K → 300K boundary,
   verifies temperature is held at 273K by latent heat absorption)

The new test runs with explicit_k(x) and checks that:

• the point nearest the hot boundary is held at 273K (int = 272)
  instead of rising freely toward 300K
• an interior point further from the boundary stays below the
  transition temperature (int = 266)

This covers the latent heat heating path in _latent_heat.py.