backend/app/services/site_finder/market_metrics.py

@@ -320,7 +320,14 @@ _SALES_WINDOW_SQL = text(
     SELECT
         COUNT(*)                          AS units_sold_window,
         COALESCE(SUM(area_pd), 0)         AS area_sold_window,
-        rooms_int
+        rooms_int,
+        -- #1214: ROLLUP grand-total и NULL-группа дают rooms_int IS NULL обе.
+        -- Различаем через GROUPING(): =1 для grand-total, =0 для NULL-комнатной
+        -- группы. Без этого один проданный лот с rooms_int IS NULL даёт ДВЕ
+        -- строки rooms_int IS NULL (NULL-группа + итог), и MixedAggregate-план
+        -- эмитит итог ПЕРВЫМ → NULL-группа затирает units_total частичным
+        -- счётом → unit_velocity/absorption занижены, MoS завышен.
+        GROUPING(rooms_int)               AS is_total
       FROM objective_lots ol
      WHERE ol.premise_kind = :premise_kind
        AND (
@@ -477,9 +484,13 @@ def _query_sales_window(
 ) -> tuple[int, float, dict[str, int]]:
     """Продажи за окно по contract_date. Возвращает (units, area_m2, {bucket: units}).
 
-    GROUP BY ROLLUP: строка с rooms_int IS NULL — это grand-total (берём как
-    units/area), остальные строки — разбивка по комнатности (для liquidity /
-    demand_concentration). На ошибке/пусто → (0, 0.0, {}).
+    GROUP BY ROLLUP с GROUPING(rooms_int) AS is_total (#1214):
+      • is_total=1 → grand-total (units/area за все комнаты);
+      • is_total=0 и rooms_int IS NULL → разбивка для лотов БЕЗ rooms — кладём
+        в by_room['unknown'] (а не путаем с total);
+      • is_total=0 и rooms_int не NULL → разбивка по конкретной комнатности.
+    by_room аккумулирует через += чтобы при будущих доп.NULL-вариантах не
+    затирать прежние счётчики. На ошибке/пусто → (0, 0.0, {}).
     """
     try:
         rows = db.execute(_SALES_WINDOW_SQL, dict(params)).mappings().all()
@@ -495,12 +506,17 @@ def _query_sales_window(
     for r in rows:
         cnt = int(r["units_sold_window"] or 0)
         area = float(r["area_sold_window"] or 0.0)
-        if r["rooms_int"] is None:
-            # ROLLUP grand-total.
+        if int(r["is_total"]) == 1:
+            # ROLLUP grand-total — единственная строка с GROUPING=1.
             units_total = cnt
             area_total = area
+        elif r["rooms_int"] is None:
+            # Лоты с rooms_int IS NULL (ETL пишет NULL для «неопределённого типа»)
+            # — отдельный бакет, не путаем с total.
+            by_room["unknown"] = by_room.get("unknown", 0) + cnt
         else:
-            by_room[_room_bucket(int(r["rooms_int"]))] = cnt
+            bucket = _room_bucket(int(r["rooms_int"]))
+            by_room[bucket] = by_room.get(bucket, 0) + cnt
     return units_total, area_total, by_room
 
 

backend/tests/services/site_finder/test_market_metrics.py

@@ -341,12 +341,13 @@ _FULL_STOCK = {
     "obj_count": 3,
     "n_long_unsold": 42,
 }
-# ROLLUP: grand-total (rooms_int=None) + per-room buckets.
+# ROLLUP с GROUPING(rooms_int) AS is_total (#1214):
+#   is_total=1 → grand-total строка; is_total=0 → разбивка по rooms_int.
 _FULL_SALES = [
-    {"units_sold_window": 60, "area_sold_window": 2880.0, "rooms_int": None},
-    {"units_sold_window": 30, "area_sold_window": 900.0, "rooms_int": 1},
-    {"units_sold_window": 20, "area_sold_window": 1200.0, "rooms_int": 2},
-    {"units_sold_window": 10, "area_sold_window": 780.0, "rooms_int": 0},
+    {"units_sold_window": 60, "area_sold_window": 2880.0, "rooms_int": None, "is_total": 1},
+    {"units_sold_window": 30, "area_sold_window": 900.0, "rooms_int": 1, "is_total": 0},
+    {"units_sold_window": 20, "area_sold_window": 1200.0, "rooms_int": 2, "is_total": 0},
+    {"units_sold_window": 10, "area_sold_window": 780.0, "rooms_int": 0, "is_total": 0},
 ]
 
 
@@ -544,7 +545,9 @@ class TestComputeMarketMetricsThinData:
             "n_long_unsold": 5,
         }
         # только grand-total строка с 0 продаж
-        sales = [{"units_sold_window": 0, "area_sold_window": 0.0, "rooms_int": None}]
+        sales = [
+            {"units_sold_window": 0, "area_sold_window": 0.0, "rooms_int": None, "is_total": 1}
+        ]
         db = _mock_db(stock, sales)
         with patch(_ELAST, return_value={"elasticity": -1.5, "source": "fallback"}):
             m = compute_market_metrics(db, district="Тихий")
@@ -622,10 +625,26 @@ class TestSalesWindowSource:
                 "units_sold_window": real_window_units,
                 "area_sold_window": 110_000.0,
                 "rooms_int": None,
+                "is_total": 1,
+            },
+            {
+                "units_sold_window": 1_000,
+                "area_sold_window": 40_000.0,
+                "rooms_int": 1,
+                "is_total": 0,
+            },
+            {
+                "units_sold_window": 800,
+                "area_sold_window": 44_000.0,
+                "rooms_int": 2,
+                "is_total": 0,
+            },
+            {
+                "units_sold_window": 508,
+                "area_sold_window": 26_000.0,
+                "rooms_int": 3,
+                "is_total": 0,
             },
-            {"units_sold_window": 1_000, "area_sold_window": 40_000.0, "rooms_int": 1},
-            {"units_sold_window": 800, "area_sold_window": 44_000.0, "rooms_int": 2},
-            {"units_sold_window": 508, "area_sold_window": 26_000.0, "rooms_int": 3},
         ]
         db = _mock_db(stock, sales)
         with patch(_ELAST, return_value={"elasticity": -1.4, "source": "regression"}):