fix(tradein): filter ДКП-only in rosreestr importer + re-enable deals #549

Merged
lekss361 merged 1 commit from feat/tradein-rosreestr-dkp-only into main 2026-05-24 19:42:29 +00:00
3 changed files with 20 additions and 11 deletions

View file

@ -834,11 +834,17 @@ async def estimate_quality(payload: TradeInEstimateInput, db: Session) -> Aggreg
except Exception as exc: except Exception as exc:
logger.warning("cian_valuation: lookup failed (graceful): %s", exc) logger.warning("cian_valuation: lookup failed (graceful): %s", exc)
# 5. Deals — отключены 2026-05-24 (см. Decision_TradeIn_DataQuality_8PR_Roadmap). # 5. Deals — ДКП-only sales (вторичка) из rosreestr_deals.
# rosreestr_deals содержит ДДУ первичного рынка (177-200 К/м²), что искажает # Importer фильтрует doc_type='ДКП' (PR-A 2026-05-24), ДДУ застройщиков
# оценку вторички (~114 К/м²). Возвращаемся к этому когда отделим вторичные # исключены — больше не скёюят median вторички ~110-120 К/м².
# сделки от ДДУ в источнике. deals = _fetch_deals(
deals: list[dict[str, Any]] = [] db,
lat=geo.lat,
lon=geo.lon,
rooms=payload.rooms,
area=payload.area_m2,
radius_m=DEFAULT_RADIUS_M,
)
# 6. Сохраняем в trade_in_estimates # 6. Сохраняем в trade_in_estimates
estimate_id = uuid4() estimate_id = uuid4()

View file

@ -102,7 +102,7 @@ def _common_patches(cian_mock):
return [ return [
patch("app.services.estimator.geocode", new=AsyncMock(return_value=_make_fake_geo())), patch("app.services.estimator.geocode", new=AsyncMock(return_value=_make_fake_geo())),
patch("app.services.estimator.get_house_metadata", new=AsyncMock(return_value=None)), patch("app.services.estimator.get_house_metadata", new=AsyncMock(return_value=None)),
patch("app.services.estimator._fetch_analogs", return_value=([], False)), patch("app.services.estimator._fetch_analogs", return_value=([], False, "W")),
patch("app.services.estimator._fetch_deals", return_value=[]), patch("app.services.estimator._fetch_deals", return_value=[]),
patch( patch(
"app.services.estimator._get_or_fetch_imv_cached", "app.services.estimator._get_or_fetch_imv_cached",
@ -134,7 +134,7 @@ def test_estimator_includes_cian_valuation_when_available() -> None:
new=AsyncMock(return_value=_make_fake_geo())), new=AsyncMock(return_value=_make_fake_geo())),
patch("app.services.estimator.get_house_metadata", patch("app.services.estimator.get_house_metadata",
new=AsyncMock(return_value=None)), new=AsyncMock(return_value=None)),
patch("app.services.estimator._fetch_analogs", return_value=([], False)), patch("app.services.estimator._fetch_analogs", return_value=([], False, "W")),
patch("app.services.estimator._fetch_deals", return_value=[]), patch("app.services.estimator._fetch_deals", return_value=[]),
patch("app.services.estimator._get_or_fetch_imv_cached", patch("app.services.estimator._get_or_fetch_imv_cached",
new=AsyncMock(return_value=None)), new=AsyncMock(return_value=None)),
@ -164,7 +164,7 @@ def test_estimator_graceful_when_cian_returns_none() -> None:
new=AsyncMock(return_value=_make_fake_geo())), new=AsyncMock(return_value=_make_fake_geo())),
patch("app.services.estimator.get_house_metadata", patch("app.services.estimator.get_house_metadata",
new=AsyncMock(return_value=None)), new=AsyncMock(return_value=None)),
patch("app.services.estimator._fetch_analogs", return_value=([], False)), patch("app.services.estimator._fetch_analogs", return_value=([], False, "W")),
patch("app.services.estimator._fetch_deals", return_value=[]), patch("app.services.estimator._fetch_deals", return_value=[]),
patch("app.services.estimator._get_or_fetch_imv_cached", patch("app.services.estimator._get_or_fetch_imv_cached",
new=AsyncMock(return_value=None)), new=AsyncMock(return_value=None)),
@ -195,7 +195,7 @@ def test_estimator_graceful_when_cian_raises() -> None:
new=AsyncMock(return_value=_make_fake_geo())), new=AsyncMock(return_value=_make_fake_geo())),
patch("app.services.estimator.get_house_metadata", patch("app.services.estimator.get_house_metadata",
new=AsyncMock(return_value=None)), new=AsyncMock(return_value=None)),
patch("app.services.estimator._fetch_analogs", return_value=([], False)), patch("app.services.estimator._fetch_analogs", return_value=([], False, "W")),
patch("app.services.estimator._fetch_deals", return_value=[]), patch("app.services.estimator._fetch_deals", return_value=[]),
patch("app.services.estimator._get_or_fetch_imv_cached", patch("app.services.estimator._get_or_fetch_imv_cached",
new=AsyncMock(return_value=None)), new=AsyncMock(return_value=None)),
@ -228,7 +228,7 @@ def test_estimator_cian_result_no_sale_price_not_added() -> None:
new=AsyncMock(return_value=_make_fake_geo())), new=AsyncMock(return_value=_make_fake_geo())),
patch("app.services.estimator.get_house_metadata", patch("app.services.estimator.get_house_metadata",
new=AsyncMock(return_value=None)), new=AsyncMock(return_value=None)),
patch("app.services.estimator._fetch_analogs", return_value=([], False)), patch("app.services.estimator._fetch_analogs", return_value=([], False, "W")),
patch("app.services.estimator._fetch_deals", return_value=[]), patch("app.services.estimator._fetch_deals", return_value=[]),
patch("app.services.estimator._get_or_fetch_imv_cached", patch("app.services.estimator._get_or_fetch_imv_cached",
new=AsyncMock(return_value=None)), new=AsyncMock(return_value=None)),

View file

@ -6,6 +6,8 @@
# rooms выводим из площади (Росреестр не отдаёт кол-во комнат). # rooms выводим из площади (Росреестр не отдаёт кол-во комнат).
# Координаты — NULL, проставляются отдельно (geocode по street). # Координаты — NULL, проставляются отдельно (geocode по street).
# #
# Только ДКП (вторичка) — ДДУ застройщиков скёюят median, отделены сознательно. См. PR-A.
#
# Запуск на прод-хосте: ./import-rosreestr.sh # Запуск на прод-хосте: ./import-rosreestr.sh
# Повторяемо: дедуп по dedup_hash, новый запуск подтянет свежие кварталы. # Повторяемо: дедуп по dedup_hash, новый запуск подтянет свежие кварталы.
@ -32,7 +34,7 @@ docker exec "$DST_PG" psql -U tradein -d tradein -v ON_ERROR_STOP=on -c "
docker exec "$SRC_PG" psql -U "$SRC_USER" -d "$SRC_DB" -v ON_ERROR_STOP=on -c " docker exec "$SRC_PG" psql -U "$SRC_USER" -d "$SRC_DB" -v ON_ERROR_STOP=on -c "
COPY ( COPY (
SELECT SELECT
md5('ros:' || id::text) AS dedup_hash, md5('ros:dkp:' || id::text) AS dedup_hash,
'Екатеринбург, ' || trim(street) AS address, 'Екатеринбург, ' || trim(street) AS address,
CASE CASE
WHEN area < 30 THEN 0 WHEN area < 44 THEN 1 WHEN area < 30 THEN 0 WHEN area < 44 THEN 1
@ -51,6 +53,7 @@ docker exec "$SRC_PG" psql -U "$SRC_USER" -d "$SRC_DB" -v ON_ERROR_STOP=on -c "
AND area BETWEEN 18 AND 200 AND area BETWEEN 18 AND 200
AND deal_price BETWEEN 1000000 AND 100000000 AND deal_price BETWEEN 1000000 AND 100000000
AND street IS NOT NULL AND trim(street) <> '' AND street IS NOT NULL AND trim(street) <> ''
AND doc_type = 'ДКП'
AND period_start_date >= '$SINCE' AND period_start_date >= '$SINCE'
) TO STDOUT WITH CSV ) TO STDOUT WITH CSV
" | docker exec -i "$DST_PG" psql -U tradein -d tradein -v ON_ERROR_STOP=on -c " " | docker exec -i "$DST_PG" psql -U tradein -d tradein -v ON_ERROR_STOP=on -c "