Revert "Use TimeSliceSelection for demand maps"

[tsmbland]

Reverts #1540

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 90.84%. Comparing base (f0c1049) to head (bd698e9).
⚠️ Report is 3 commits behind head on main.

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #1546   +/-   ##
=======================================
  Coverage   90.83%   90.84%           
=======================================
  Files          61       61           
  Lines        9014     9019    +5     
  Branches     9014     9019    +5     
=======================================
+ Hits         8188     8193    +5     
  Misses        507      507           
  Partials      319      319           

☔ 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.

[Copilot]

🟡 Changes recommended

A critical unmet-demand tolerance issue and missing regression coverage remain unresolved.

Once you've addressed the issues Copilot identified, you can request another Copilot review.

Pull request overview

Reverts #1540, restoring time-slice-keyed demand handling across investment appraisal and market selection.

Changes:

• Restores per-time-slice demand flattening and appraisal handling.
• Updates constraints, output, fixtures, benchmarks, and regression data.
• Restores prior time-slice serialization behavior.

File summaries

File	Summary
tests/data/simple/debug_appraisal_results.csv	Updates regression expectations.
src/time_slice.rs	Restores prior enum serialization behavior.
src/simulation/market.rs	Uses time-slice demand lookups.
src/simulation/investment/appraisal/optimisation.rs	Computes per-time-slice unmet demand.
src/simulation/investment/appraisal/constraints.rs	Aggregates demand for constraints.
src/simulation/investment.rs	Restores time-slice demand maps and flattening.
src/output.rs	Restores per-slice appraisal output fields.
src/fixture.rs	Updates appraisal fixtures.
benches/assets.rs	Updates benchmark demand handling.

Review details

Suppressed comments (2)

src/simulation/investment.rs:163

• This conversion is the only path that turns annual/seasonal input demand into the per-time-slice map consumed by appraisal, but the unit tests in this module only exercise already-flattened maps. Please add parameterised coverage for at least annual and seasonal selections, asserting that each selection's per-slice values sum to the original demand and that separate selections do not overwrite one another; otherwise a regression here can silently change investment demand.

            // We split the demand equally over all time slices in the selection
            // NOTE: since demands will only be balanced to the time slice level of the commodity
            // it doesn't matter how we do this distribution, only the total matters.
            #[allow(clippy::cast_precision_loss)]
            let n_time_slices = time_slice_selection.iter(time_slice_info).count() as f64;
            let demand_per_slice = *demand / Dimensionless(n_time_slices);
            for (time_slice, _) in time_slice_selection.iter(time_slice_info) {
                demand_map.insert(
                    (commodity_id.clone(), region_id.clone(), time_slice.clone()),
                    demand_per_slice,
                );
            }

src/simulation/investment/appraisal/constraints.rs:56

• This flow lookup is now inside the time-slice loop even though the coefficient is constant for the whole selection. Appraisal runs this constraint construction for every candidate and round, so this adds one IndexMap lookup per time slice unnecessarily; hoist the lookup before the inner loop.

            let flow_coeff = asset.get_flow(&commodity.id).unwrap().coeff;

• Files reviewed: 9/10 changed files
• Comments generated: 1
• Review effort level: Lite

---

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