106 commits

Author	SHA1	Message	Date
Light1YT	714d4fd3be	feat(#978b): survivorship-free Source A + detrend control in rate-sensitivity backtest ... Cleaner OOS validation of the §9.6 negative verdict: - --source {B,A,both}: Source A = objective_corpus_room_month SUM(deals) monthly aggregate (survivorship-FREE, ~13mo); Source B path unchanged. - --detrend: linearly detrend ln(units) before differencing so a spurious survivorship trend can't drive β (_detrend_log helper). - cross-source verdict compares B-raw / B-detrended / A. DEFINITIVE RESULT (run vs prod 2026-06-03): NO variant beats coin-flip OOS — B-raw 0.148, B-detrended 0.148, A too thin (13mo<18). The §9.6 negative verdict is a REAL 'no signal', NOT a survivorship artifact (detrend + clean Source A agree). monthly key_rate→sales is not OOS-predictive on available EKB data → §9.6 stays advisory; forecast should lean on measured signals (supply/absorption). Tests 38→60.	2026-06-03 14:51:43 +05:00
bot-backend	6f06ca7f16	feat(forecasting): read-only OOS backtest of §9.6 rate-sensitivity (#978) (#1024)	2026-06-03 09:33:02 +00:00
bot-backend	12162a0abb	feat(exporters): Site Finder report PDF exporter (#989, 955-A3) (#1023)	2026-06-03 09:09:30 +00:00
bot-backend	299f2365fb	feat(exporters): Excel exporter for SiteFinderReport (#991, 955-A5) (#1022)	2026-06-03 08:57:00 +00:00
bot-backend	fc312fe606	feat(forecasting): §13 report assembler (#988, 955-A2) (#1021)	2026-06-03 08:51:23 +00:00
bot-backend	47fe5a55eb	feat(forecasting): §15 confidence engine v2 (#990, 955-A4) (#1020)	2026-06-03 08:41:07 +00:00
bot-backend	75c9b34987	feat(forecasting): §13 SiteFinderReport object (#987, 955-A1) (#1019)	2026-06-03 08:33:01 +00:00
bot-backend	008f9d9ed1	feat(forecasting): §25 six special indices (#986, 954-C) (#1018)	2026-06-03 08:24:08 +00:00
bot-backend	98dea0315a	feat(forecasting): §14.2 product scoring card (#985, 954-B) (#1017)	2026-06-03 08:11:54 +00:00
bot-backend	8057468c13	feat(forecasting): §11 macro-scenarios (#984, 954-A) (#1016)	2026-06-03 07:59:22 +00:00
bot-backend	691ccef4b7	feat(forecasting): class/commercial/USP §10.2/10.4/10.5 + §16 (#983, 953-B) (#1015)	2026-06-03 07:52:24 +00:00
bot-backend	72e9b24f2c	feat(recommend): horizon-aware recommend_mix opt-in overlay (#982, 953-A) (#1014)	2026-06-03 07:40:33 +00:00
bot-backend	489e380f1e	feat(forecasting): what-to-build ranker + MAI proxy (#981, 952-B) (#1013)	2026-06-03 07:07:17 +00:00
bot-backend	70ffa399fc	feat(forecasting): demand-supply forecast engine (#980, 952-A) (#1012)	2026-06-03 06:56:33 +00:00
bot-backend	9758e21cbd	feat(forecasting): §9.4 demand-normalization coefficient (#951f, advisory) (#1011)	2026-06-03 06:28:14 +00:00
bot-backend	69ad6c87fc	feat(forecasting): §9.5 macro coefficient (#951e, advisory) (#1010)	2026-06-03 06:20:04 +00:00
bot-backend	f41e2536d8	feat(forecasting): §9.6 key-rate sensitivity module (#951d, advisory) (#1009)	2026-06-03 06:06:16 +00:00
bot-backend	c7bfc9e22a	feat(forecasting): monthly sales series builder for §9.6 (#951c) (#1008)	2026-06-03 05:52:33 +00:00
bot-backend	ace3b99508	feat(forecasting): monthly macro series + regime classifier (#951b) (#1007)	2026-06-03 05:37:43 +00:00
lekss361	a10592847d	feat(site_finder): future-supply-pressure index (#950 Step 6) (#1006)	2026-06-03 05:10:08 +00:00
Light1YT	ae85f19993	feat(workers): supply-layers refresh task + weekly beat (#950 Step 5) (#1005)	2026-06-02 20:47:05 +00:00
Light1YT	900802264a	feat(site-finder): supply_layers v2 compute service (#950 EPIC6 step3+4) (#1004)	2026-06-02 20:34:26 +00:00
Light1YT	59be55f80e	feat(site-finder): per-competitor relevance_weight (#949 PR B) (#1000)	2026-06-02 19:45:38 +00:00
Light1YT	45d61ecff0	feat(site_finder): market-metrics service (#949 PR A, §9.2) (#997)	2026-06-02 19:26:26 +00:00
Light1YT	c78c6ec2d0	feat(site-finder): live CBR key-rate scraper → macro_indicator (#945 PR B) (#964)	2026-06-02 19:05:57 +00:00
Light1YT	dbae4b0bda	feat(site-finder): macro_indicator table + backfill + reader (#945 PR A) (#963)	2026-06-02 18:41:28 +00:00
bot-backend	aa8ca9bcfc	feat(backend): add /api/v1/ping liveness endpoint (#634) ... Co-authored-by: bot-backend <bot-backend@gendsgn.local> Co-committed-by: bot-backend <bot-backend@gendsgn.local>	2026-05-29 15:53:29 +00:00
Light1YT	6cb3d7515c	fix(tradein): per-bucket velocity formula + realistic срок продажи (#574) (#631)	2026-05-28 15:35:16 +00:00
Light1YT	07106d744a	chore(mv): regression test + sql rule for weighted AVG (#295) (#626)	2026-05-28 13:49:54 +00:00
Light1YT	c08e4d7811	feat(rbac): pilot scope → только /trade-in/** (#587) ... Decision 2026-05-26: pilot не видит landing/analytics/site-finder/concept, только Trade-In.	2026-05-26 07:40:44 +00:00
Light1YT	88cdfd6adb	feat(rbac): role-based access control via X-Authenticated-User middleware (#585) ... Backend RBAC + Caddy header_up. Closes part of #fixes-rbac-pra.	2026-05-26 06:18:40 +00:00
lekss361	f9adb5cc2f	feat(22d): catalog scraper — skip-today filter + force flag (#348)	2026-05-17 23:15:22 +00:00
lekss361	0567ad2130	feat(22d): domrf_catalog_object scraper — fill ~25 NULL kn_objects cols from SSR __NEXT_DATA__ (#335)	2026-05-17 21:26:22 +00:00
lekss361	c61c8c86af	feat(sf-b1): GET /parcels/by-bbox — map entry endpoint + parcel_user_status table (#336)	2026-05-17 21:25:48 +00:00
lekss361	b013fd886c	feat(sf-b6): GET /parcels/{cad}/poi-score — weighted top-7 (#333)	2026-05-17 21:11:25 +00:00
lekss361	7a8d59beb4	feat(obj-3): SF backend migrate to objective_lots ground truth (price 0.3% -> 81%) (#324)	2026-05-17 20:08:17 +00:00
lekss361	689327909c	fix(sf-09): capacity-aware redistribute (round 2) — bounded + correct invariant ... Round 1 (commit bcd7dc8) был broken: на 2-bucket входах surplus уходил в free полностью без учёта capacity → free превышал cap → следующая итерация clamp'ировала его и наоборот. Infinite oscillation в FastAPI handler. Round 2 fix per review BLOCK (#282 comment): - Surplus распределяется пропорционально available capacity (cap - v), не текущему v. Free никогда не вылетит выше cap. - free = строго < cap (не <=) — иначе деление на 0 capacity. - Hard guard `for _ in range(N+1)` — гарантированно завершается. - Pathological (surplus > total_capacity): возвращаем оригинальный pct_map + cap_skipped=True (sum=100 invariant сохранён). - Hamilton round вынесен в _hamilton_round() helper. Tests: - 2-bucket cases (90/10, 70/30, 99/1) expected cap_skipped=True - test_cap_iteration_count_bounded — все pathological завершаются < 100ms - All 13 cases verified standalone (3 fast-path + 7 reproduced + 3 pathological)	2026-05-17 15:30:50 +03:00
lekss361	bcd7dc8f75	fix(sf-09): iterative cap algorithm + cap_skipped flag + unit tests ... Review verdict MINOR (PR #282) — single-pass cap_and_redistribute не гарантировал invariant max ≤ MAX_BUCKET_SHARE_PCT. Surplus от clamped поднимал free выше cap (например {1k:75, studio:15, 2k:10} → studio=39). - backend/app/services/site_finder/best_layouts.py - _cap_and_redistribute → iterative while-loop до сходимости - Returns (result_map, cap_skipped) — флаг для pathological all-clamped - Float work map, Hamilton финальный pass для invariant sum=100 - backend/app/schemas/parcel.py - LayoutTzRecommendation.cap_skipped: bool = False - backend/tests/services/site_finder/test_best_layouts.py - 6 parametrized tests: invariants, 7 failing cases from review, no-dominant unchanged, empty, cap_skipped propagation, normal-case - frontend/src/types/best-layouts.ts — cap_skipped: boolean - frontend/src/components/site-finder/BestLayoutsBlock.tsx - banner условие сменилось с maxPct > 60 на rec.cap_skipped (honest signal)	2026-05-17 15:21:51 +03:00
lekss361	1799e1fb30	fix(sf-02): добавить site_status/ready_dt в /analyze competitors + sort строящиеся first ... SQL в /analyze не выбирал site_status, ready_dt из domrf_kn_objects; ORDER BY flat_count DESC ставил сданные ЖК в топ. Closes (epic part) #271 item 2	2026-05-17 13:29:59 +03:00
lekss361	c701636515	Merge forgejo/main into fix/sf-01-velocity-time-window ... Resolve conflict in best_layouts.py: keep SF-01 inline velocity (deals_window numerator) + add SF-05 sold_pct clamp at 100% and is_oversold flag from main.	2026-05-17 12:15:40 +03:00
lekss361	5a4408e0ab	fix(sf-01): time_window honest velocity — inline SQL с реальным фильтром report_month ... Раньше _VELOCITY_DIVISORS делил агрегаты mv_layout_velocity (24 мес) на 4/12 для quarter/year, не меняя реальное окно данных. Теперь inline SQL из objective_corpus_room_month с CAST(:window_interval AS interval). velocity_per_month = deals_window / months_in_window (1.0/3.0/12.0). Разные time_window → разные строки из БД → разный mix/velocity/jk_count. Closes (epic part) #271 item 1	2026-05-17 12:12:19 +03:00
lekss361	247c252b04	fix(sf-05): clamp sold_pct_of_supply at 100% + add is_oversold flag ... Symptom: sold_pct_of_supply = sum_deals_24mo / supply_count_snapshot * 100 yields >100% (e.g. 199%) for fast-selling small formats due to incompatible time windows (24-month deals vs point-in-time supply snapshot). Option A (hotfix): clamp to 100.0, expose is_oversold=True for UI badge. - best_layouts.py: compute sold_pct_raw, clamp with min(..., 100.0), set is_oversold = raw > 100% - parcel.py TopLayoutRow: add is_oversold: bool field - best-layouts.ts TopLayoutRow: add is_oversold: boolean - BestLayoutsBlock.tsx: show warn badge ">100%" when is_oversold=True - tests: two new cases — raw 199% clamps + is_oversold=True; raw 50% passes through + is_oversold=False Closes (epic part) #271 item 5	2026-05-17 11:57:29 +03:00
lekss361	f160c5efaf	Merge pull request 'feat(#263 sub-A): cad_* overlay schemas + cadastre territorial_zones (ПЗЗ)' (#265) from feat/263-sub-a-cad-nspd-overlay-schemas-and-pzz into main	2026-05-17 07:50:08 +00:00
lekss361	8e377059a4	fix(#265 review): CRS 3857→4326 + stable zone_id hash + rename 102 + tx comment ... - #1 (BLOCK): ST_GeogFromGeoJSON → ST_Transform(ST_SetSRID(..., 3857), 4326)::geography so EPSG:3857 features from get_territorial_zones_in_bbox are correctly stored in GEOGRAPHY(POLYGON, 4326) column (matches cad_parcels/cad_buildings pattern) - #2: zone_id fallback changed from counter-based f"{quarter_cad}_{inserted}" to md5(json(props))[:12] — stable between runs, prevents wrong-row ON CONFLICT - #3: rename 101_cad_nspd_overlay_tables.sql → 102_* (101_user_custom_pois.sql exists) - #4: add comment clarifying SQLAlchemy autobegin guarantees outer tx before begin_nested() - Tests: updated test_feature_without_zone_id_uses_fallback to assert md5 format + idempotency	2026-05-17 10:45:52 +03:00
lekss361	f08db43145	feat(#263 sub-A): cad_* overlay schemas + cadastre territorial_zones (ПЗЗ) ... Phase 1 (migration 101): создаёт 5 новых таблиц cad_territorial_zones, cad_red_lines, cad_engineering_structures, cad_risk_zones, cad_opportunity_parcels. Idempotent (IF NOT EXISTS), GIST-индексы. Phase 2 partial (ПЗЗ only): - NSPDBulkClient.get_features_in_bbox_grid — async grid-walk WMS GetFeatureInfo - NSPDBulkClient.get_territorial_zones_in_bbox — convenience wrapper layer 875838 - harvest_quarter Phase 2.5 — вызов get_territorial_zones_in_bbox после grid-walk - _save_territorial_zones — UPSERT loop с SAVEPOINT per-row, ON CONFLICT zone_id 9 unit tests (mock DB): insert/upsert/no-geom/no-id/error-isolation/savepoint pattern.	2026-05-17 10:28:22 +03:00
lekss361	26ee3b015c	Merge pull request 'fix(#261): custom_pois — regression tests для db.commit() в create/update/delete (P0 persistence)' (#262) from fix/261-custom-pois-missing-commit into main	2026-05-17 07:26:54 +00:00
lekss361	8ebcd15ed6	Merge pull request 'fix(#231): switch search_by_quarter area layers to grid-walk (Sub-PR B)' (#260) from fix/231-nspd-search-by-quarter-grid into main	2026-05-17 07:26:08 +00:00
lekss361	570c987e78	fix(#231): repair test_rate_limit_semaphore regression + docstring sync ... Re-review feedback (PR #260 head 8191a85): 1. tests/scrapers/test_nspd_bulk_client.py:360 — test referenced bulk_mod._SEMAPHORE which was renamed to _SEMAPHORE_LIMIT in previous commit. AttributeError на полном pytest run. Заменено: assert bulk_mod._SEMAPHORE_LIMIT == 3 (smoke). max_concurrent <= 3 уже валидирует throttling — capacity check был лишним. 2. app/services/scrapers/nspd_client.py docstrings (lines 483, 506) — обновлены ссылки NSPDBulkClient._SEMAPHORE(3) → NSPDBulkClient._sem (per-instance, capacity=3) чтобы соответствовать новой архитектуре.	2026-05-17 10:22:16 +03:00
lekss361	c94545f0eb	fix(#261): add db.commit() regression tests для custom_pois service ... Проблема: POST /custom-pois возвращал 201 с id=N, но row не сохранялась в БД — user_custom_pois оставалась пустой после request close. Симптом возникает если service-функция не вызывает db.commit() явно: get_db.finally → db.close() без commit → psycopg3 rollbacks pending tx. Код (#257) содержит db.commit() во всех 3 мутациях, но тесты полностью мокировали service layer через patch() — regression в commit не ловился. Fix: добавлены 6 service-level тестов с паттерном "два сеанса": - MagicMock(db) + db.commit.assert_called_once() для create/update/delete - db.commit.assert_not_called() для not-found и empty-payload случаев Closes #261	2026-05-17 10:16:36 +03:00
lekss361	8191a85102	fix(#231): address re-review blockers — per-instance semaphore + time_limit + test mocks ... PR #260 second-opinion review 🔴 BLOCK → fixes: 1. nspd_bulk_client.py: _SEMAPHORE module-level → per-instance self._sem создаётся в __aenter__ под текущий event loop. Sequential asyncio.run() в search_by_quarter (11 layers × asyncio.run() per layer) bind'ил module-level semaphore к first loop'у → RuntimeError на втором вызове → все cells return_exceptions=True проглатывали → grid возвращал 0 features. 2. nspd_sync.py: harvest_quarter добавлен time_limit=900 (hard) рядом с soft_time_limit=600. Safety net если task игнорит SoftTimeLimitExceeded. 3. tests/test_nspd_client.py: добавлен _make_fake_grid_walk helper + monkeypatch NSPDClient.get_features_in_bbox_grid в 3 тестах (search_by_quarter_core_only / _with_zouit / _layers_fetched_with_risks). Без этого area layers били бы по живому NSPD API → CI flaky. Tests: 37 passed, 1 skipped (3 area-dispatch + 34 grid-walk).	2026-05-17 10:13:34 +03:00