feat(workflow-operator): score the rows an empty cell leaves usable

[kz930]

What changes were proposed in this PR?

The Machine Learning Scorer handed both configured columns straight to scikit-learn's metric functions, so one empty cell reached them as a None and ended the run with "Classification metrics can't handle a mix of unknown and binary targets". The message named neither the column nor a way out, and no configuration could exclude the row.

The operator now drops the rows missing either value before scoring, the way the visualization operators drop missing values before plotting. The subset names both columns rather than dropping on each separately: dropping on either alone would leave y_true and y_pred misaligned.

Any related issues, documentation, discussions?

Closes #8058

How was this PR tested?

A new case in MachineLearningScorerOpDescSpec pins the drop in the generated code. Run by hand on a table with a blank in the actual-value column and a blank in the predicted-value column: before, the run ended with the ValueError above; after, it reports Accuracy 0.6667 over the three scorable rows.

Was this PR authored or co-authored using generative AI tooling?

Generated-by: Claude Code (Claude Opus 5)

[kz930]

/request-review @aglinxinyuan @carloea2

[github-actions]

Automated Reviewer Suggestions

Based on the git blame history of the changed files, we recommend the following reviewers:

• Contributors with relevant context: @aglinxinyuan
  You can notify them by mentioning @aglinxinyuan in a comment.

[codecov-commenter]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 93.42%. Comparing base (eddec26) to head (9e68d4d).
⚠️ Report is 8 commits behind head on main.

Additional details and impacted files

@@             Coverage Diff              @@
##               main    #8070      +/-   ##
============================================
- Coverage     93.43%   93.42%   -0.02%     
+ Complexity     4710     4707       -3     
============================================
  Files          1181     1181              
  Lines         47801    47810       +9     
  Branches       5326     5327       +1     
============================================
+ Hits          44663    44666       +3     
- Misses         1698     1700       +2     
- Partials       1440     1444       +4     

Flag	Coverage Δ		*Carryforward flag
access-control-service	81.00% <ø> (ø)
agent-service	99.32% <ø> (ø)		Carriedforward from eddec26
amber	89.58% <100.00%> (-0.03%)	⬇️
computing-unit-managing-service	73.67% <ø> (ø)
config-service	86.73% <ø> (ø)
file-service	87.89% <ø> (ø)
frontend	95.90% <ø> (ø)		Carriedforward from eddec26
notebook-migration-service	79.31% <ø> (ø)
pyamber	97.74% <ø> (ø)		Carriedforward from eddec26
workflow-compiling-service	77.19% <ø> (ø)

*This pull request uses carry forward flags. Click here to find out more.

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

🚀 New features to boost your workflow:

• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[github-actions]

⚠️ Benchmark changes need a look

🟢 4 better · 🔴 1 worse · ⚪ 10 noise (<±5%) · 0 without baseline

Compared against main bdc6d2a benchmarked on this same runner, so the delta is largely free of cross-runner hardware noise. The "7d avg" column still reflects the gh-pages dashboard. Treat <±5% as noise unless repeated.

Dashboard · Run

	config	throughput	MB/s	latency	max Δ latest / 7d
🟢	bs=10 sw=10 sl=64	407	0.249	24,642/30,836/30,836 us	🟢 -8.0% / 🔴 +120.5%
🔴	bs=100 sw=10 sl=64	895	0.546	109,836/133,462/133,462 us	🔴 +6.4% / 🔴 +40.3%
⚪	bs=1000 sw=10 sl=64	1,053	0.643	946,312/1,006,072/1,006,072 us	⚪ within ±5% / 🔴 -11.3%

Baseline details

Latest main bdc6d2a from same runner

config	metric	PR	latest main	7d avg	Δ latest	Δ 7d
bs=10 sw=10 sl=64	throughput	407 tuples/sec	414 tuples/sec	893.99 tuples/sec	-1.7%	-54.5%
bs=10 sw=10 sl=64	MB/s	0.249 MB/s	0.252 MB/s	0.546 MB/s	-1.2%	-54.4%
bs=10 sw=10 sl=64	p50	24,642 us	23,599 us	11,176 us	+4.4%	+120.5%
bs=10 sw=10 sl=64	p95	30,836 us	33,521 us	14,009 us	-8.0%	+120.1%
bs=10 sw=10 sl=64	p99	30,836 us	33,521 us	16,898 us	-8.0%	+82.5%
bs=100 sw=10 sl=64	throughput	895 tuples/sec	925 tuples/sec	1,157 tuples/sec	-3.2%	-22.7%
bs=100 sw=10 sl=64	MB/s	0.546 MB/s	0.565 MB/s	0.706 MB/s	-3.4%	-22.7%
bs=100 sw=10 sl=64	p50	109,836 us	103,218 us	88,763 us	+6.4%	+23.7%
bs=100 sw=10 sl=64	p95	133,462 us	141,293 us	95,096 us	-5.5%	+40.3%
bs=100 sw=10 sl=64	p99	133,462 us	141,293 us	102,854 us	-5.5%	+29.8%
bs=1000 sw=10 sl=64	throughput	1,053 tuples/sec	1,040 tuples/sec	1,187 tuples/sec	+1.3%	-11.3%
bs=1000 sw=10 sl=64	MB/s	0.643 MB/s	0.635 MB/s	0.724 MB/s	+1.3%	-11.2%
bs=1000 sw=10 sl=64	p50	946,312 us	957,307 us	874,968 us	-1.1%	+8.2%
bs=1000 sw=10 sl=64	p95	1,006,072 us	1,026,390 us	917,815 us	-2.0%	+9.6%
bs=1000 sw=10 sl=64	p99	1,006,072 us	1,026,390 us	949,868 us	-2.0%	+5.9%

Raw CSV

config_idx,batch_size,schema_width,string_len,num_batches,total_ms,total_tuples,total_bytes,tuples_per_sec,mb_per_sec,lat_p50_us,lat_p95_us,lat_p99_us
0,10,10,64,20,490.84,200,128000,407,0.249,24641.77,30836.10,30836.10
1,100,10,64,20,2235.25,2000,1280000,895,0.546,109835.96,133461.96,133461.96
2,1000,10,64,20,18994.71,20000,12800000,1053,0.643,946311.67,1006072.10,1006072.10