fix: Repair data pipeline with StatCan CMHC rental data

- Add StatCan CMHC parser to fetch rental data from Statistics Canada API
- Create year spine (2014-2025) as time dimension driver instead of census
- Add CMA-level rental and income intermediate models
- Update mart_neighbourhood_overview to use rental years as base
- Fix neighbourhood_service queries to match dbt schema
- Add CMHC data loading to pipeline script

Data now flows correctly: 158 neighbourhoods × 12 years = 1,896 records
Rent data available 2019-2025, crime data 2014-2024

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

portfolio_app/toronto/loaders/__init__.py

@@ -3,7 +3,12 @@
 from .amenities import load_amenities, load_amenity_counts
 from .base import bulk_insert, get_session, upsert_by_key
 from .census import load_census_data
-from .cmhc import load_cmhc_record, load_cmhc_rentals
+from .cmhc import (
+    ensure_toronto_cma_zone,
+    load_cmhc_record,
+    load_cmhc_rentals,
+    load_statcan_cmhc_data,
+)
 from .cmhc_crosswalk import (
     build_cmhc_neighbourhood_crosswalk,
     disaggregate_zone_value,
@@ -32,6 +37,8 @@ __all__ = [
     # Fact loaders
     "load_cmhc_rentals",
     "load_cmhc_record",
+    "load_statcan_cmhc_data",
+    "ensure_toronto_cma_zone",
     # Phase 3 loaders
     "load_census_data",
     "load_crime_data",

portfolio_app/toronto/loaders/cmhc.py

@@ -1,5 +1,9 @@
 """Loader for CMHC rental data into fact_rentals."""
 
+import logging
+from datetime import date
+from typing import Any
+
 from sqlalchemy.orm import Session
 
 from portfolio_app.toronto.models import DimCMHCZone, DimTime, FactRentals
@@ -8,6 +12,12 @@ from portfolio_app.toronto.schemas import CMHCAnnualSurvey, CMHCRentalRecord
 from .base import get_session, upsert_by_key
 from .dimensions import generate_date_key
 
+logger = logging.getLogger(__name__)
+
+# Toronto CMA zone code for CMA-level data
+TORONTO_CMA_ZONE_CODE = "TORCMA"
+TORONTO_CMA_ZONE_NAME = "Toronto CMA"
+
 
 def load_cmhc_rentals(
     survey: CMHCAnnualSurvey,
@@ -135,3 +145,117 @@ def load_cmhc_record(
         return _load(session)
     with get_session() as sess:
         return _load(sess)
+
+
+def ensure_toronto_cma_zone(session: Session | None = None) -> int:
+    """Ensure Toronto CMA zone exists in dim_cmhc_zone.
+
+    Creates the zone if it doesn't exist.
+
+    Args:
+        session: Optional existing session.
+
+    Returns:
+        The zone_key for Toronto CMA.
+    """
+
+    def _ensure(sess: Session) -> int:
+        zone = (
+            sess.query(DimCMHCZone).filter_by(zone_code=TORONTO_CMA_ZONE_CODE).first()
+        )
+        if zone:
+            return int(zone.zone_key)
+
+        # Create new zone
+        new_zone = DimCMHCZone(
+            zone_code=TORONTO_CMA_ZONE_CODE,
+            zone_name=TORONTO_CMA_ZONE_NAME,
+            geometry=None,  # CMA-level doesn't need geometry
+        )
+        sess.add(new_zone)
+        sess.flush()
+        logger.info(f"Created Toronto CMA zone with zone_key={new_zone.zone_key}")
+        return int(new_zone.zone_key)
+
+    if session:
+        return _ensure(session)
+    with get_session() as sess:
+        result = _ensure(sess)
+        sess.commit()
+        return result
+
+
+def load_statcan_cmhc_data(
+    records: list[Any],  # List of CMHCRentalRecord from statcan_cmhc parser
+    session: Session | None = None,
+) -> int:
+    """Load CMHC rental data from StatCan parser into fact_rentals.
+
+    This function handles CMA-level data from the StatCan API, which provides
+    aggregate Toronto data rather than zone-level HMIP data.
+
+    Args:
+        records: List of CMHCRentalRecord dataclass instances from statcan_cmhc parser.
+        session: Optional existing session.
+
+    Returns:
+        Number of records loaded.
+    """
+    from portfolio_app.toronto.parsers.statcan_cmhc import (
+        CMHCRentalRecord as StatCanRecord,
+    )
+
+    def _load(sess: Session) -> int:
+        # Ensure Toronto CMA zone exists
+        zone_key = ensure_toronto_cma_zone(sess)
+
+        loaded = 0
+        for record in records:
+            if not isinstance(record, StatCanRecord):
+                logger.warning(f"Skipping invalid record type: {type(record)}")
+                continue
+
+            # Generate date key for this record's survey date
+            survey_date = date(record.year, record.month, 1)
+            date_key = generate_date_key(survey_date)
+
+            # Verify time dimension exists
+            time_dim = sess.query(DimTime).filter_by(date_key=date_key).first()
+            if not time_dim:
+                logger.warning(
+                    f"Time dimension not found for {survey_date}, skipping record"
+                )
+                continue
+
+            # Create fact record
+            fact = FactRentals(
+                date_key=date_key,
+                zone_key=zone_key,
+                bedroom_type=record.bedroom_type,
+                universe=record.universe,
+                avg_rent=float(record.avg_rent) if record.avg_rent else None,
+                median_rent=None,  # StatCan doesn't provide median
+                vacancy_rate=float(record.vacancy_rate)
+                if record.vacancy_rate
+                else None,
+                availability_rate=None,
+                turnover_rate=None,
+                rent_change_pct=None,
+                reliability_code=None,
+            )
+
+            # Upsert
+            inserted, updated = upsert_by_key(
+                sess, FactRentals, [fact], ["date_key", "zone_key", "bedroom_type"]
+            )
+            loaded += inserted + updated
+
+        logger.info(f"Loaded {loaded} CMHC rental records from StatCan")
+        return loaded
+
+    if session:
+        return _load(session)
+    with get_session() as sess:
+        result = _load(sess)
+        sess.commit()
+        return result