Write vel(2) to 2D probe files and drop the per-step print

[sbryngelson]

Fixes #1841.

For a plain 2D case (no bubbles, no hypoelasticity), s_write_probe_files wrote only (time, rho, vel(1), pres) — the second velocity component was dropped — and it also printed a line to stdout on every time step for every probe:

write (i + 30, '(6X,F12.6,F24.8,F24.8,F24.8)') nondim_time, rho, vel(1), pres
print *, 'time =', nondim_time, 'rho =', rho, 'pres =', pres

The 3D branch writes all three components and the 2D bubble and hypoelastic branches both write vel(2), so this looks like the plain-2D branch being incomplete rather than intentional. The print floods the log of any long run: one line per probe per step.

The fypp guard matches the neighbouring branches, since vel(2) does not exist under 1D case optimization.

Heads up for reviewers: this changes the 2D probe file from four columns to five, so any 2D probe golden files need regenerating. I have not run a 2D probe case with this patch — I hit the missing component while post-processing and fixed it, but my later 2D runs used no probes and my 3D runs take a different branch. Worth a quick check before merge.

./mfc.sh precheck passes.

https://claude.ai/code/session_01HMJ7cycfo7kTFSFq5yhHLG

[Copilot]

Warning

Copilot couldn't run its full agentic review because it didn't start before the timeout. Make sure your repository has a runner available, or add a copilot-code-review.yml file specifying one with the runs-on attribute. See the docs for more details.

Pull request overview

This PR fixes incomplete probe output for plain 2D simulations by writing the second velocity component (vel(2)) to probe files and removing a per-step print that floods stdout/logs.

Changes:

• Update the plain-2D probe output to include vel(2) (expanding output from 4 to 5 columns).
• Remove per-probe, per-timestep stdout logging from s_write_probe_files.
• Add a fypp guard to avoid referencing vel(2) under 1D case optimization.

---

💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.

[github-actions]

Claude Code Review

Head SHA: aa71e84

Files changed:

• 1
• src/simulation/m_data_output.fpp

Findings:

• src/simulation/m_data_output.fpp (runtime-info else branch, ~line 1593): the new write is wrapped in #:if not MFC_CASE_OPTIMIZATION or num_dims > 1 with no num_dims == 1 fallback, so under --case-optimization with num_dims == 1 this entire write/output statement is compiled away — the file previously always wrote a line here (nondim_time, rho, vel(1), pres), and now nothing is written at all for 1D case-optimized builds. This also creates a case-optimization-dependent behavior divergence: a non-case-optimized build with num_dims == 1 still hits this line (since not MFC_CASE_OPTIMIZATION is true) and writes vel(2) for a 1D run, while a case-optimized num_dims == 1 build silently drops the write entirely — the same physical case produces different diagnostic output (or none) purely based on the case-optimization compile flag. The fix should preserve the original 4-field write (vel(1) only) for num_dims == 1 and only use the new 5-field vel(1), vel(2) write when num_dims > 1, rather than eliding the branch outright.

[codecov]

Codecov Report

❌ Patch coverage is 0% with 1 line in your changes missing coverage. Please review.
✅ Project coverage is 61.26%. Comparing base (dc0aec1) to head (0cf529a).

Files with missing lines	Patch %	Lines
src/simulation/m_data_output.fpp	0.00%	1 Missing ⚠️

Additional details and impacted files

@@           Coverage Diff           @@
##           master    #1844   +/-   ##
=======================================
  Coverage   61.26%   61.26%           
=======================================
  Files          84       84           
  Lines       22330    22329    -1     
  Branches     3265     3265           
=======================================
  Hits        13680    13680           
+ Misses       6207     6206    -1     
  Partials     2443     2443           

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

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

[sbryngelson]

Now exercised at runtime, which I had flagged above as missing.

2D case (flat plate in a freestream, 3 probes, --gpu mp, 4 ranks on Frontier). The probe file has five columns where it previously had four, and the second velocity component reads correctly:

# t                rho             u               v               p
0.000833       1.00000000      1.00000000      0.00000000     17.85714286

u = 1, v = 0 and p = 17.857 are the prescribed freestream, so the new column is both present and right. The stdout flood is gone.

Reminder for whoever merges: 2D probe golden files need regenerating, since the file goes from four columns to five.

[github-actions]

Lines of Code

File	Lines	Diff
src/simulation/m_data_output.fpp	1368	+1

Directory	Lines	Diff
simulation	28087	+1
total	46341	+1