Captures the grill-with-docs session for solid-wall insulation: CONTEXT.md
gains EWI/IWI Measure Types + the Wall Insulation Eligibility rule (+ a
flagged-ambiguity that the three planning flags stay distinct, never recollapsed
to legacy restricted_measures). ADR-0019 records the eligibility policy (cavity
-> cavity only; brick/system -> IWI+EWI; timber -> IWI only; cob/stone -> none;
conservation/flat block EWI, listed/heritage block both). ADR-0020 records
conservation/listed/heritage as three distinct Property attributes sourced by
extending the geospatial S3 repo (flags co-located with lat/long).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Slice 1 of solid-wall insulation. BuildingPartOverlay gains a
wall_insulation_thickness field; the generic applicator already folds it onto
SapBuildingPart by name. With wall_insulation_type=1 (EWI) / 3 (IWI) + 100 mm,
the calculator derives the post-insulation U-value (§5.8 documentary path,
λ=0.04 default) — and for IWI also lowers the thermal-mass parameter. Two new
Elmhurst before/after cascade pins (solid-brick EWI + IWI, cert 001431)
reproduce the re-lodged after at abs(diff) <= 1e-4 across SAP/CO2/PE.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
fetch_epc_bulk_sample streams certificates-<year>.json out of the bulk ZIP via
range requests, keeps the first N SAP-version matches, and writes each cert's
inner document to <out>/<cert>.json for run_property_report. Stops after N, so
only the member prefix transfers, not the 15.7 GB archive (RangeFile.bytes_read
reports the true transfer vs the absolute ZIP offset). Verified on 2026: 100
SAP-10.2 certs -> report ran 81 scorable (MAE 2.03), 46 flagged, 19 raises
(11 full-SAP schema 19.1.0, 7 unmapped floor_construction 0/3, 1 missing
post_town) — real shadow-validation signal vs the curated golden 57.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The bulk endpoint 302-redirects to a 15.7 GB S3 ZIP with one NDJSON member per
year; each line wraps the per-cert payload in a stringified 'document' that
parses to the same RdSAP-Schema-21.0.1 shape from_api_response already handles.
parse_bulk_line unwraps a record; is_sap_version filters to SAP 10.2; RangeFile
exposes the S3 object as a seekable file so zipfile streams a single year's
member (and a sampler stops early) without downloading the whole archive.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
run_property_report builds the three-section Markdown+CSV report over a dir of
API-shaped EPC JSON, offline (defaults to the golden 57: 57/57 scorable, MAE
0.54, 6 flagged |Δ|>0.5). fetch_epc_dump pulls raw cert JSON from the live API
by --uprn/--postcode (picking the latest cert per match, skipping existing
files), mirroring fetch_cohort2's proven HTTP shape and reading
OPEN_EPC_API_TOKEN. Report artifacts + epc_dump/ are gitignored.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
format_report_csv emits one comma-safe row per property: the calculator-error
fields (lodged/calculated/Δ/flag), the Plan headline figures (baseline+post
SAP/band, measures, cost+contingency, bill & CO2 savings, valuation %), the
flattened measure triggers, and any captured error — sortable in a spreadsheet
for a large dump.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
format_report_markdown emits: (1) cohort parity stats + a per-property
lodged-vs-calculated table flagging |Δ| > 0.5 (errors shown inline),
(2) Plans + costings (SAP/band jump, cost + contingency, bill & CO2 savings,
valuation uplift), (3) each fired measure with the EPC attributes that
triggered it.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
build_property_reports models a dump in order (errors captured per-cert);
parity_report_for aggregates the lodged-vs-calculated SAP across the cohort
into the existing ParityReport (MAE/RMSE/bias/worst-N), excluding certs that
couldn't be mapped or scored. Residual convention is the calculator's own
(predicted - actual), the negative of PropertyReport.sap_error.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Section 3 of the report: build_property_report now runs the Modelling stage
and, for every Plan Measure, records the EPC attribute(s) that caused its
generator to fire (MeasureTrigger) — wall_construction/insulation for cavity
fill, roof thickness for loft, floor thickness/construction for floors, the
absent mechanical kind for ventilation. Modelling raises are captured as
plan_error, independent of the calculator-error capture.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Section 1 of the property inspection report: PropertyReport compares the
cert's lodged energy_rating_current to Sap10Calculator's un-rounded SAP and
flags |Δ| > 0.5 (the ADR-0010/0013 shadow-validation design target). A
mapping/scoring raise is captured per-cert as calculator_error, never
propagated, so one bad cert can't abort the sweep.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Next-phase handover: fetch live EPCs via EpcClientService, run the
offline Modelling harness, and save a per-property report covering
(1) lodged-vs-calculated SAP divergence (>0.5), (2) plans + costings,
(3) recommended measures + the EPC attributes that triggered them. Maps
the EPC API client (the user's blocker), the calculator-error ingredients
(parity_report scaffolding), and each generator's exact trigger fields.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Apply the deep-research off-gas figures so oil/smokeless/wood sit on the
same NEP-Apr-2026 retail / DESNZ DUKES gross-CV basis as the new coal
proxy (fuel-input, not useful-heat): OIL 9.16 -> 12.11 (prior value was
materially low vs current kerosene), SMOKELESS 10.0 -> 8.69, WOOD_LOGS
8.83 -> 8.25, WOOD_PELLETS 7.99 -> 7.38. SEG (15.0, Solar Energy UK) and
LPG (17.61, bottled-propane) kept; gas/electricity (Ofgem cap) unchanged.
CV arithmetic recorded in the snapshot _assumptions. OIL pin updated.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Coal and heat networks have no national retail/cap rate, so the snapshot
left them null and BillDerivation raised UnpricedFuel — dropping those
certs from an offline cohort run. Add researched proxy rates (fuel-input
basis, sources + arithmetic in the JSON _note/_gaps): COAL 7.13 p/kWh
(NEP Nov 2025 coal uprated + DESNZ DUKES house-coal GCV) and HEAT_NETWORK
16.0 p/kWh + 69.4 p/day (Insite Energy operator sample; indicative, schemes
vary ~8-30). Both flagged proxy/indicative — sense-check estimates, not
market rates. Existing curated fuels are unchanged.
Replaces the unpriced-raises pin for these two with a positive rate pin;
off-peak stays unpriced pending the day/night accessor. Golden cohort now
runs 57/57 offline with zero errors.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
CertResult now carries its Plan (with flat baseline/post-SAP/measures
properties), and `format_cohort_csv` renders one browsable row per cert
(SAP transition, band, measures, cost, bill saving, valuation %, error).
`scripts/run_modelling_cohort.py` is turnkey: no args runs the committed
golden cohort, prints a sense-check table for the first measure-bearing
certs (a capped preview so a large dump doesn't flood the terminal), the
summary, and writes modelling_cohort.csv (gitignored). Point it at the
EPC dump when it lands.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
`harness.cohort.run_cohort(paths)` parses each API-shaped EPC JSON with
from_api_response and models it via run_modelling — no database, no
network — capturing per-cert errors instead of aborting the sweep, plus
`format_cohort_summary`. A thin `scripts/run_modelling_cohort.py` CLI
points it at a directory. Proven over the 57 golden API certs: 56 ran
offline, 15 produced measures, 1 errored (COAL has no Fuel Rates entry —
a BillDerivation coverage gap, not a harness one). Ready for the EPC dump.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Two fixes that unblock offline, no-database inspection over an arbitrary
EPC dump:
- Complete the harness sample catalogue with loft_insulation and
solid_floor_insulation — the four fabric generators can emit five
Measure Types, but the catalogue priced only three, so an offline run
on a property with an uninsulated loft or solid floor raised mid-run.
A new test pins the catalogue to cover every generator Measure Type.
- Add `run_modelling(epc, ...)` — runs ONLY the Modelling stage (no
Ingestion / Baseline), so it needs no lodged recorded-performance / RHI
and inspects recommendations on any calculator-scorable EPC. `run_one`
(full pipeline) stays for when you want Baseline too.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The Plan derives its Valuation Uplift (ADR-0018) from its baseline -> post
band jump and works+contingency cost, given one external input — the
Property's current market value (a Property Valuation, mostly absent).
`Plan.valuation` / `Plan.baseline_epc_rating` are derived like the other
headline figures; `PlanModel.from_domain` maps the £ forms to the live
plan.valuation_* columns (NULL when no value — the percentage is not
persisted on those columns). `Property.current_market_value` is the new
optional source; the orchestrator threads it onto the Plan. `run_one`
takes a `current_market_value` so the harness can value the uplift, and
the sense-check table shows the average % (always) plus the £ forms when
known.
Sourcing the current market value (upload / default) remains deferred
(ADR-0018); it is None throughout until that lands, so the columns stay
NULL at scale.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The financial-uplift model per ADR-0018. `estimate_valuation_uplift(
current_band, target_band, current_value=None, total_cost=None)` returns
a `ValuationUplift`: band-transition uplift compounded from four broker
tables (MoneySupermarket / Lloyds per-step, Knight Frank / Rightmove
whole-jump), taking min/max/mean across the covering sources. Always a
percentage; absolute £ forms (increase at each bound + post-retrofit
value) only when a current market value is supplied; the 2x ROI cap
rescales the percentages and can only bite once a value is known. A
non-improving jump is a clean 0% no-op.
Pure function, no external dependency. Persisting it (where the value
lands) and sourcing the current market value stay deferred (ADR-0018).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Expand half of the recommendation_materials retirement (ADR-0017). A
Plan Measure installs a single Product, so thread its catalogue id end to
end — Product.id -> MeasureOption.material_id -> PlanMeasure.material_id
-> recommendation.material_id — replacing the per-material BOM child
table with one nullable column on the row. ProductPostgresRepository
reads the id from MaterialRow; the four fabric generators set it on their
Option; the orchestrator carries it onto the Plan Measure; the mirror
declares + maps the column. Optional throughout (the JSON stopgap
catalogue carries no ids -> NULL).
The multi-measure integration test now pins each persisted measure's
material_id to its seeded MaterialRow id. Migration spec (live column
must be added before this deploys; contraction is the owner's next step)
in docs/migrations/recommendation-material-id.md.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Slice 3. `harness.console.run_one(epc, goal_band=...)` wires the full
AraFirstRunPipeline against in-memory fakes — no Postgres, no network —
runs one property, prints the sense-check table, and returns the Plan
for interactive poking from a REPL at the worktree root. Defaults to the
committed harness sample catalogue.
Refactors the slice-1 integration test to delegate to run_one (dropping
~70 lines of duplicated wiring + the now-unused test catalogue fixture),
so it exercises the shipped entrypoint rather than a parallel copy. The
new console test covers run_one's print/return contract.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Slice 2. `harness.plan_table.format_plan_table(plan)` renders a Plan as a
plain-text table — one package summary line (baseline SAP/band -> post
SAP/band, CO2 saved, cost of works + contingency, bill saved) and one
line per Plan Measure (signed SAP points, cost, delivered kWh + £
savings). Pure presentation: reads the Plan, computes nothing. The
DB-less First Run test now prints it (visible under `pytest -s`) so the
modelled package can be eyeballed and debugged by hand.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
From the grill-with-docs pass on the depth+scale phase. Splits the
overloaded "valuation" into two glossary terms — Property Valuation
(current market value, a Baseline attribute, mostly missing) and
Valuation Uplift (plan-conditional, percentage-primary; absolute £ only
when a Property Valuation exists, 2x ROI cap on the £ form). ADR-0018
records the percentage-primary decision and why (the EPC scale corpus
has no market values, so a value-primary model produces nothing), plus
the deferred sourcing / per-measure / rental-yield items.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Slice 1 of the DB-less inspection harness. Complete the in-memory
FakeUnitOfWork so the ModellingOrchestrator runs with no Postgres:
add FakeScenarioRepository + FakePlanRepository (idempotent, keyed by
(property_id, scenario_id)), expose scenario/product/plan on the fake
unit, and grow FakePropertyRepo to compose the effective EPC from the
EPC repo at read time — mirroring PropertyPostgresRepository, so the
EPC Ingestion persists is visible to Baseline + Modelling (the
through-repos hand-off, in memory).
The new integration test drives the full AraFirstRunPipeline
(Ingestion -> Baseline -> Modelling) against the FakeUnitOfWork — no
Session ever opened — on the uninsulated 000490 fixture with its lodged
recorded-performance filled in (it already carries the RHI block, so
Baseline can run) and asserts a multi-measure Plan is produced. The
committed product catalogue prices the wall/floor/ventilation measures
it fires.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Capture the next phase (close persisted-field gaps + financial uplift, plus a
large-scale e2e run of a SAP 10.2 EPC dump and console manual testing; measure
coverage deferred) and a self-contained handover prompt for a fresh agent to
pick up via a grilling session.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Pin the FE-facing aggregate_portfolio_recommendations (previously untested): it
sums a Scenario's default Recommendations onto the Scenario row, joining
Recommendation → Plan on recommendation.plan_id. Locks the m2m→plan_id read cut
for the FE-critical path, now testable thanks to the full-parity ScenarioModel.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Move the scenario and installed_measure tables into
infrastructure/postgres/modelling/ as full-parity SQLModel definitions
(ScenarioModel, InstalledMeasureModel + MeasureType), completing the cluster
consolidation. backend/app/db/models/recommendations.py is now a pure
re-export shim.
ScenarioModel.goal is the PortfolioGoal enum (legacy planning branches on it),
sourced from domain/modelling/portfolio_goal.py; the repo's to_domain maps it to
its value string, so domain Scenario.goal is now the value ("Increasing EPC")
consistent with the orchestrator's check — fixing the latent name-vs-value
inconsistency the old str column masked (the scenario repo test stored the enum
*name*). Parity columns are nullable (mirror convention; live NOT-NULLs owned by
Drizzle).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
PortfolioGoal is domain vocabulary (a Scenario's goal — legacy planning branches
on PortfolioGoal.INCREASING_EPC), so it belongs in domain/ co-located with
scenario.py, mirroring how domain/epc/wall_type.py holds an enum that
infrastructure/ imports. This lets the consolidated ScenarioModel (next slice)
source the goal enum from domain without an infra→backend dependency.
portfolio.py keeps a re-export so every existing
`from ...portfolio import PortfolioGoal` caller is unaffected.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Standardise the modelling persistence classes on the …Model suffix (PlanModel,
RecommendationModel, RecommendationMaterialModel) — matching the epc_property
precedent and the legacy names the rest of backend/ already imports, so the
shim's plan re-export becomes literal (no alias) and the eventual shim deletion
needs zero renames. The …Row→…Model sweep for the non-cluster tables
(Property/Task/Material/…) waits until their live legacy …Model counterparts
are retired, to avoid reintroducing dual-definition collisions. No behaviour
change.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Stop writing the m2m (remove create_plan_recommendations + its call, the bulk
link insert and the now-dead plan_ids_by_index, and the plan_recommendations
delete in delete_property_batch) and remove the PlanRecommendationRow model +
its shim alias and the test_export fixture inserts. Measures now link to their
Plan solely via recommendation.plan_id (writers set it, readers join on it).
The live drop of the plan_recommendations table is the FE-owned Drizzle
migration documented in docs/migrations/recommendation-plan-id.md, sequenced
after the read-cut + backfill.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Rewrite the three structurally-identical m2m-join readers
(portfolio_functions.aggregate_portfolio_recommendations,
Outputs.get_recommendations_from_db, export get_recommendations) to join
PlanModel directly via recommendation.plan_id, dropping the plan_recommendations
join and its now-unused import. The writers set plan_id (prior slice), so the
rows resolve. test_export pins the export reader through the cut (its fixtures
now set recommendation.plan_id). A portfolio_functions DB characterization test
lands with the scenario consolidation (which provides the full-parity scenario
table the aggregation writes to).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
upload_recommendations and bulk_upload_recommendations_and_materials now set
plan_id on each recommendation row (the plan id is already in scope), while
still writing the plan_recommendations m2m — the dual-write that lets readers
move onto plan_id with no breakage during the transition (ADR-0017 amendment /
docs/migrations/recommendation-plan-id.md). The m2m write is removed in a later
slice once no reader depends on it.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Move the live plan, recommendation, recommendation_materials and (retiring)
plan_recommendations tables into a new infrastructure/postgres/modelling/
subpackage as single SQLModel definitions (the epc_property pattern), absorbing
the rebuild's partial PlanRow/RecommendationRow mirrors and carrying full
legacy column parity plus recommendation.plan_id. Out-of-cluster references are
plain indexed ints (mirror convention); the live FKs are owned by the Drizzle
schema. backend/app/db/models/recommendations.py becomes a re-export shim
(ScenarioModel/InstalledMeasure stay for a later slice).
Fix the export conftest to create SQLModel-first (so Base funding_package's FK
to the now-SQLModel plan resolves) and skip the redundant drop_all on its
function-scoped throwaway DB (the epc enum type is now shared across both
metadatas). Resolves the pre-existing dual-definition collision: the rebuild
and legacy export suites are now co-runnable. No behaviour change.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Rewrite the migration spec into the full expand/contract sequence (add plan_id
→ backfill → dual-write → cut reads → drop) with the two load-bearing rules:
backfill before any read cuts over, and dual-write the m2m until all reads are
off it (the Drizzle FE reads the tables directly, so the repos can't deploy
atomically). Amend ADR-0017 from "m2m retired for new writes" to "m2m dropped +
one SQLModel definition per table under infrastructure/postgres/modelling/".
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
`_plan_for` now scores the baseline + every cumulative prefix once
(`cascade_scores`, best-practice order) and reuses those Scores for both the
role-3 marginal attribution and a per-measure bill cascade: bill each prefix at
one Fuel Rates snapshot and take consecutive Bill deltas as each measure's
marginal delivered-kWh and £ saving. Saving is signed (ventilation is
negative) and telescopes exactly to the Plan headline savings, because the
Plan's baseline/post Bills are now the same cascade endpoints (`bills[0]` /
`bills[-1]`) — which also drops the redundant standalone baseline `calculate`.
`recommendation.kwh_savings` / `energy_cost_savings` are filled from these.
Adds `Bill.total_consumption_kwh` (shared by Plan + the orchestrator). Pinned
end-to-end on the real calculator: Σ per-measure savings == the Plan totals
(ADR-0014 amendment).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
`PlanMeasure` grows optional `kwh_savings` (delivered energy) and
`energy_cost_savings` (£) — its slice of the telescoping bill cascade, signed
so positive is a saving and `None` until billing runs. `RecommendationRow`
declares the matching live `recommendation.kwh_savings` /
`energy_cost_savings` columns and maps them in `from_domain` (None → NULL).
The vestigial `recommendation.energy_savings` stays undeclared (legacy = 0).
No FE migration — the columns already exist on the live table (ADR-0014 / 0017).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Pull the cumulative-prefix scoring out of `marginal_impacts` into a reusable
`cascade_scores(scorer, baseline, overlays) -> list[Score]` (index 0 the
baseline, one calculator run per prefix) plus a pure `marginals_from_scores`.
Each Score carries its SapResult, so the next slice's telescoping per-measure
bill cascade can re-bill the same prefixes the role-3 attribution already
scores — no extra `calculate` calls (ADR-0014 / ADR-0016). `marginal_impacts`
now delegates; behaviour unchanged.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
ModellingOrchestrator gains a constructor-injected FuelRatesRepository (mirrors
Baseline): run() resolves get_current() once and reuses one BillDerivation across
the batch. _plan_for prices the baseline and post-package end-states from the
SapResults already on their Scores (no extra calculate) and passes the Bills to
Plan. PlanRow mirror + from_domain gain the four live columns post_energy_bill /
energy_bill_savings / post_energy_consumption / energy_consumption_savings.
Pipeline/handler wire the fuel-rates repo. Integration tests assert the columns
persist: the multi-measure (fallback) plan shows positive bill+consumption
savings; the already-at-target zero-measure plan shows the current bill with
exactly zero savings. Fuel-switch measures price at the new fuel for free (we
bill the simulated end-state). 183 modelling/billing/orchestration/repo tests
pass, pyright strict clean. Plan-level only; per-measure savings next.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Plan gains optional baseline_bill / post_bill (the Bills derived for the
unmodified and post-package end-states at one Fuel Rates snapshot) and derives
the four plan-level columns: post_energy_bill (post total), energy_bill_savings
(baseline - post), post_energy_consumption (Σ post section kWh), and
energy_consumption_savings (baseline - post delivered kWh). All return None until
billing runs (persisted as NULL), so existing Plan construction and the
not-yet-wired orchestrator stay green. Plan-level only; per-measure savings are a
later slice (ADR-0014 amendment).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Score gains sap_result: Optional[SapResult], populated by PackageScorer with the
calculator output its headline figures came from. This lets the Modelling stage
price the post-package (and baseline) end-state via Bill Derivation reusing a
SapResult already computed by the optimiser's re-score / the orchestrator's
baseline score — no second calculate (ADR-0014 amendment). The optimiser reads
only sap_continuous, so it stays domain-agnostic and the stub scorers (which omit
sap_result) are unaffected — all optimiser tests pass unchanged.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Bill / EnergyBreakdown / BillDerivation / sap_fuel were under
domain/property_baseline/ only because Baseline was built first. The Modelling
stage now needs them too, so move them (and their tests) to a neutral
domain/billing/ — Fuel/FuelRates already live in the shared domain/fuel_rates/.
Avoids a modelling -> property_baseline cross-stage import and a package name
that wrongly implies ownership (ADR-0011, ADR-0014 amendment). Pure git mv +
import rewrite across 10 files; 40 billing/baseline/repo tests pass, pyright
strict clean. CONTEXT.md Bill Derivation location updated.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Records the /grill-with-docs design for the Modelling Bill-Derivation slice:
Bill Derivation is cross-stage (relocate Bill/EnergyBreakdown/BillDerivation/
sap_fuel to a neutral domain/billing/); Modelling bills the fully-overlaid
post-package SapResult (so fuel-switch measures price at the new fuel for free),
diffing against the baseline at the same FuelRates snapshot; the post-package
and baseline SapResults are captured from scores the optimiser/orchestrator
already compute (Score.sap_result), so no second calculate; FuelRatesRepository
is constructor-injected into ModellingOrchestrator mirroring Baseline; plan-level
columns this slice, per-measure telescoping bill cascade next (energy_savings is
vestigial, left NULL).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The orchestrator already threads budget/target_sap/dependencies into
optimise_package, so no orchestrator change was needed. Add an integration test
proving the new objective end-to-end on the real calculator: a band-D property
(~57.4) with a goal of band D — already met — yields a Plan with NO measures and
zero cost (the old max-gain objective would have recommended wall+floor+vent,
improving within the band it is already in). Clarified that the existing
multi-measure test now exercises the max-gain fallback (goal C unreachable from
D, tops out ~61). Narrowed Optional sap_points/estimated_cost through locals to
keep pyright strict-clean.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The warm-start (and max-gain fallback) now price each forced Measure Dependency
the candidate triggers, not just inject it afterwards: optimise/optimise_min_cost
fold dependencies into each candidate's cost+gain via _augmented_cost_gain, and
optimise_package scores each dependency's true role-1 signal (_with_role1_signals)
instead of the 0.0 placeholder. This stops the min-cost objective (i) ignoring the
~£900 a wall drags in (a wall-free package reaching target can be cheaper) and
(ii) picking a small-gain wall whose mandatory ventilation (down to -5 SAP) makes
it net-negative, which repair cannot un-pick.
Budget is now a hard envelope: the constraint applies to the augmented (measure +
its ventilation) cost, so a wall that fits alone but whose ventilation would bust
the budget is DROPPED rather than forced over budget. This reverses the earlier
'forced regardless of budget' call (which made sense when selection was
ventilation-blind). Safety invariant intact — presence still injected on every
path; we just never recommend a wall we can't afford to ventilate. ADR-0016
amendment updated. 94 modelling+orchestration tests pass.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Rewire the objective per the ADR-0016 amendment. With a target_sap (Increasing
EPC): warm-start optimise_min_cost (cheapest package reaching target_gain =
target_sap - baseline within budget) -> inject dependencies -> re-score ->
repair toward target; if the warm-start is infeasible or the repaired package
still falls short on the true score, fall back to max-gain-within-budget (best
effort). Without a target_sap: max-gain (unchanged). The min-cost objective
stops at the target without overshooting into a higher band; surplus budget is
left unspent. Extracted _max_gain_package (no-target path + fallback) and
_repair_to_target (inject + re-score + greedy repair). Dependency injection and
the repair loop are preserved; all prior optimiser + dependency tests pass
unchanged. Ventilation-aware *selection* is the next slice; injection is still
post-warm-start here.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Exact-enumeration sibling to optimise(): pick <=1 option per group to minimise
total cost subject to total gain >= target_gain and cost <= budget (None =
unconstrained). Ties broken toward higher gain ('recommend more'). Returns None
when no package within budget reaches the target (caller falls back to
max-gain); a non-positive target is met by the empty package. This is the
warm-start objective for an Increasing EPC goal per the ADR-0016 amendment
(least-cost-to-target, not max-gain). Dependency-blind for now; ventilation-aware
selection lands in a later slice.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The original ADR-0016 mis-specified the warm-start objective as maximise-gain-
subject-to-budget (with the target a repair floor); the rebuild faithfully
implemented that wrong objective. The intended behaviour is the legacy
StrategicOptimiser Case 1: minimise cost subject to (true) SAP gain >= target and
cost <= budget, falling back to max-gain-within-budget only when the target is
unreachable. For Increasing EPC this is least-cost-to-target: cheapest package
reaching the band, stops at the target (no overshoot into a higher band), surplus
budget unspent.
Also records: target predicate sap_continuous >= band floor (conservative, no
legacy slack — re-score+repair supersede it); ventilation-aware selection (the
forced dependency, -1 to -5 SAP, is folded into candidate evaluation with a real
negative role-1 signal, not just injected afterwards); presence-vs-awareness
enforcement; warm-start+re-score+repair structure and scalability rationale kept.
Sharpened the CONTEXT.md Optimised Package definition to match.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
