docs: correct feature count (2,514 actual, not 3,043 claimed)

Fixed documentation mismatch where code claimed 3,043 features but
actual implementation uses 2,514 features:
- Dataset reality: 2,553 columns = 1 timestamp + 38 targets + 2,514 features
- Breakdown: 603 full-horizon + 12 partial + 1,899 historical
- Context shape: (504, 2517) = 504h × (ts + border + target + 2,514 features)

Updated documentation in:
- src/forecasting/dynamic_forecast.py (header + docstring)
- src/forecasting/chronos_inference.py (header + print statements)
- app.py (UI description)

This ensures documentation matches code reality for accurate handover.

app.py

@@ -2,7 +2,7 @@
 """
 FBMC Chronos-2 Forecasting API
 HuggingFace Space Gradio Interface
-Version: 1.4.0 (past-only masking with 3,043 features)
+Version: 1.4.0 (past-only masking with 2,514 features)
 """
 
 # CRITICAL: Set PyTorch memory allocator config BEFORE any imports
@@ -140,9 +140,9 @@ with gr.Blocks(title="FBMC Chronos-2 Forecasting") as demo:
     generalizes directly to FBMC cross-border flows using historical patterns and future covariates.
 
     **Features**:
-    - 3,043 engineered features using past-only covariate masking
+    - 2,514 engineered features using past-only covariate masking
     - Known-future: weather, generation, load forecasts (615 features)
-    - Past-only masked: CNEC outages, volatility, flows (~2,428 features)
+    - Past-only masked: CNEC outages, volatility, flows (1,899 features)
     - 24-month historical context (Oct 2023 - Oct 2025)
     - Time-aware extraction (prevents data leakage)
     - Probabilistic forecasts (9 quantiles: 1st/5th/10th/25th/50th/75th/90th/95th/99th)

src/forecasting/chronos_inference.py

@@ -2,7 +2,7 @@
 """
 Chronos-2 Inference Pipeline with Past-Only Covariate Masking
 Standalone inference script for HuggingFace Space deployment.
-Uses predict_df() API with ALL 3,043 features leveraging Chronos-2's mask-based attention.
+Uses predict_df() API with ALL 2,514 features leveraging Chronos-2's mask-based attention.
 FORCE REBUILD: v1.4.0 - Past-only covariates + batch_size=128 for volatility capture
 """
 
@@ -173,10 +173,10 @@ class ChronosInferencePipeline:
         total_start = time.time()
 
         # PER-BORDER INFERENCE WITH PAST-ONLY COVARIATE MASKING
-        # Using predict_df() API with ALL 3,043 features (known-future + past-only masked)
-        print(f"\n[PAST-ONLY MASKING] Running inference for {len(forecast_borders)} borders with 3,043 features...")
+        # Using predict_df() API with ALL 2,514 features (known-future + past-only masked)
+        print(f"\n[PAST-ONLY MASKING] Running inference for {len(forecast_borders)} borders with 2,514 features...")
         print(f"  Known-future: weather, generation, load forecasts (615 features)")
-        print(f"  Past-only masked: CNEC outages, volatility, historical flows (~2,428 features)")
+        print(f"  Past-only masked: CNEC outages, volatility, historical flows (1,899 features)")
 
         for i, border in enumerate(forecast_borders, 1):
             # Clear GPU cache BEFORE each border to prevent memory accumulation

src/forecasting/dynamic_forecast.py

@@ -9,10 +9,10 @@ Key Concepts:
 - run_date: When the forecast is made (e.g., "2025-09-30 23:00")
 - forecast_horizon: Always 14 days (D+1 to D+14, fixed at 336 hours)
 - context_window: Historical data before run_date (typically 512 hours)
-- future_covariates: ALL 3,043 features (leveraging Chronos-2 past-only masking)
+- future_covariates: ALL 2,514 features (leveraging Chronos-2 past-only masking)
   * 603 full-horizon (known future)
   * 12 partial D+1 (masked D+2-D+14)
-  * ~2,428 historical (masked as past-only covariates)
+  * 1,899 historical (masked as past-only covariates)
 
 Chronos-2 Past-Only Covariate Masking:
 - Historical features have NaN future values → Chronos-2 sets mask=0
@@ -170,10 +170,10 @@ class DynamicForecast:
         """
         Extract future covariate data for D+1 to D+14.
 
-        Future covariates include ALL 3,043 features using Chronos-2's past-only masking:
+        Future covariates include ALL 2,514 features using Chronos-2's past-only masking:
         - Full-horizon D+14: 603 features (known future values)
         - Partial D+1: 12 features (load forecasts, masked D+2-D+14)
-        - Historical: ~2,428 features (MASKED as past-only covariates)
+        - Historical: 1,899 features (MASKED as past-only covariates)
 
         Past-only covariates leverage Chronos-2's mask-based attention:
         - Future values are NaN (unknown)