hoàn thành chức năng predict ndvi time series analysis

api_server.py

@@ -4230,6 +4230,26 @@ async def ndvi_predict_timeseries(config: NDVIPredictionConfig):
         print(f"  Mean NDVI: {result['mean_ndvi']:.3f}")
         print(f"  NDVI range: [{result['min_ndvi']:.3f}, {result['max_ndvi']:.3f}]")
         print(f"{'='*70}\n")
+
+        # Auto-save prediction cache (Sentinel-2 data + metadata) similar to rice predict
+        try:
+            cache_config = {
+                "min_lon": config.bbox[0],
+                "min_lat": config.bbox[1],
+                "max_lon": config.bbox[2],
+                "max_lat": config.bbox[3],
+                "start_date": config.start_date,
+                "end_date": config.end_date,
+                "max_scenes": config.max_scenes,
+                "cloud_cover": config.max_cloud_cover,
+                "resolution": config.resolution,
+                "model_filename": config.model_filename
+            }
+
+            cache_result = save_prediction_cache_sync(cache_config, s2_data)
+            print(f"[NDVI TIMESERIES][CACHE] Saved cache: {cache_result.get('filename')} (with data: {s2_data is not None})")
+        except Exception as cache_exc:
+            print(f"[NDVI TIMESERIES][CACHE ERROR] {cache_exc}")
         
         return result
         
@@ -4380,10 +4400,17 @@ async def ndvi_forecast(config: NDVIForecastConfig):
         
         # Mask cloud pixels
         scl = s2_data['SCL']
-        cloud_mask = ~np.isin(scl, [3, 8, 9, 10, 0, 1])
+        # cloud_mask: shape (time, y, x) if scl is xarray.DataArray, else (time, y, x) ndarray
+        # If scl is xarray.DataArray, cloud_mask will be xarray.DataArray, else numpy ndarray
+        if hasattr(scl, 'isel'):
+            cloud_mask = ~np.isin(scl, [3, 8, 9, 10, 0, 1])  # xarray.DataArray
+        else:
+            # scl is numpy ndarray, so cloud_mask is ndarray
+            cloud_mask = ~np.isin(scl, [3, 8, 9, 10, 0, 1])
         
         # LAND-TYPE-SPECIFIC FORECASTING
         use_ml_classification = config.model_filename is not None
+        force_simple_forecast = False
         
         if use_ml_classification:
             print(f"\n🤖 Using ML model for land-type-specific forecasting...")
@@ -4447,10 +4474,8 @@ async def ndvi_forecast(config: NDVIForecastConfig):
                     continue
             
             if len(point_features) == 0:
-                raise HTTPException(
+                print("[NDVI FORECAST][FALLBACK] No valid classification points. Falling back to simple seasonal forecast.")
-                    status_code=404, 
+                force_simple_forecast = True
-                    detail=f"Không tìm thấy điểm hợp lệ để phân loại trong khu vực này. Thử: (1) Chọn khu vực lớn hơn, (2) Tăng historical_months, (3) Giảm max_cloud_cover, hoặc (4) Chọn khu vực có dữ liệu vệ tinh tốt hơn. Đã thử {config.sample_points} điểm ngẫu nhiên nhưng tất cả đều bị masked (mây/nước)."
-                )
             
             print(f"✅ Extracted features for {len(point_features)} valid points")
             
@@ -4480,8 +4505,13 @@ async def ndvi_forecast(config: NDVIForecastConfig):
                 month = time_val.month
                 
                 # Get valid pixels for this time step
-                mask_t = cloud_mask.isel(time=time_idx)
+                if hasattr(cloud_mask, 'isel'):
-                
+                    mask_t = cloud_mask.isel(time=time_idx)
+                else:
+                    # Wrap numpy mask to xarray with same coords/dims as ndvi slice
+                    ndvi_slice = ndvi.isel(time=time_idx)
+                    mask_t = xr.DataArray(mask_t := cloud_mask[time_idx], coords=ndvi_slice.coords, dims=ndvi_slice.dims)
+
                 ndvi_t = ndvi.isel(time=time_idx).where(mask_t)
                 ndwi_t = ndwi.isel(time=time_idx).where(mask_t)
                 ndbi_t = ndbi.isel(time=time_idx).where(mask_t)
@@ -4539,55 +4569,67 @@ async def ndvi_forecast(config: NDVIForecastConfig):
                         }
             
             print(f"✅ Calculated patterns for {len(land_type_seasonal_stats)} land types")
+
+            if len(land_type_seasonal_stats) == 0:
+                print("[NDVI FORECAST][FALLBACK] No seasonal patterns. Falling back to simple seasonal forecast.")
+                force_simple_forecast = True
             
             # Generate forecast using land-type-weighted average
-            print(f"\n🔮 Generating land-type-specific forecast...")
+            if force_simple_forecast:
-            
+                print("[NDVI FORECAST][FALLBACK] Skipping ML forecast, will use simple seasonal averaging.")
-            forecast_timeseries = []
+            else:
-            current_date = forecast_start
+                print(f"\n🔮 Generating land-type-specific forecast...")
-            
-            # Calculate land type weights
-            total_points = len(predictions)
-            land_type_weights = {lt: np.sum(predictions == lt) / total_points 
-                                for lt in unique_types}
-            
-            while current_date <= forecast_end:
-                month = current_date.month
                 
-                # Aggregate forecast across all land types (weighted)
+                forecast_timeseries = []
-                weighted_forecast = {
+                current_date = forecast_start
-                    'ndvi_mean': 0, 'ndvi_min': 0, 'ndvi_max': 0, 'ndvi_std': 0,
-                    'ndvi_range': 0, 'ndwi_mean': 0, 'ndbi_mean': 0, 'evi_mean': 0
-                }
                 
-                land_type_contributions = {}
+                # Calculate land type weights
+                total_points = len(predictions)
+                if total_points == 0:
+                    print("[NDVI FORECAST][FALLBACK] Zero valid classified points. Switching to simple seasonal forecast.")
+                    force_simple_forecast = True
+                else:
+                    land_type_weights = {lt: np.sum(predictions == lt) / total_points 
+                                        for lt in unique_types}
                 
-                for land_type, weight in land_type_weights.items():
+                if not force_simple_forecast:
-                    if land_type in land_type_seasonal_stats and month in land_type_seasonal_stats[land_type]:
+                    while current_date <= forecast_end:
-                        stats = land_type_seasonal_stats[land_type][month]
+                        month = current_date.month
                         
-                        land_type_contributions[int(land_type)] = {
+                        # Aggregate forecast across all land types (weighted)
-                            **stats,
+                        weighted_forecast = {
-                            'weight': float(weight)
+                            'ndvi_mean': 0, 'ndvi_min': 0, 'ndvi_max': 0, 'ndvi_std': 0,
+                            'ndvi_range': 0, 'ndwi_mean': 0, 'ndbi_mean': 0, 'evi_mean': 0
                         }
                         
-                        for key in weighted_forecast:
+                        land_type_contributions = {}
-                            weighted_forecast[key] += stats[key] * weight
+                        
-                
+                        for land_type, weight in land_type_weights.items():
-                if land_type_contributions:
+                            if land_type in land_type_seasonal_stats and month in land_type_seasonal_stats[land_type]:
-                    forecast_data = {
+                                stats = land_type_seasonal_stats[land_type][month]
-                        'date': current_date.strftime('%Y-%m-%d'),
+                                
-                        'is_forecast': True,
+                                land_type_contributions[int(land_type)] = {
-                        'land_type_specific': land_type_contributions,
+                                    **stats,
-                        **weighted_forecast
+                                    'weight': float(weight)
-                    }
+                                }
-                    forecast_timeseries.append(forecast_data)
+                                
-                
+                                for key in weighted_forecast:
-                current_date += relativedelta(months=1)
+                                    weighted_forecast[key] += stats[key] * weight
-            
+                        
-            method_used = "Land-Type-Specific Forecasting (ML-Enhanced)"
+                        if land_type_contributions:
-            
+                            forecast_data = {
-        else:
+                                'date': current_date.strftime('%Y-%m-%d'),
+                                'is_forecast': True,
+                                'land_type_specific': land_type_contributions,
+                                **weighted_forecast
+                            }
+                            forecast_timeseries.append(forecast_data)
+                        
+                        current_date += relativedelta(months=1)
+                    
+                    method_used = "Land-Type-Specific Forecasting (ML-Enhanced)"
+
+        if force_simple_forecast or not use_ml_classification:
             # Simple seasonal averaging (fallback)
             print(f"\n📈 Calculating simple seasonal patterns (no ML)...")
             

ndvi_interface.html

@@ -3,7 +3,7 @@
 <head>
     <meta charset="UTF-8">
     <meta name="viewport" content="width=device-width, initial-scale=1.0">
-    <title>NDVI Time Series Analysis</title>
+    <title>NDVI Time Series Predicts</title>
     
     <!-- Leaflet CSS -->
     <link rel="stylesheet" href="https://unpkg.com/leaflet@1.9.4/dist/leaflet.css" />
@@ -131,6 +131,22 @@
             opacity: 0.5;
             cursor: not-allowed;
         }
+
+        .region-filter-btn {
+            padding: 3px 10px;
+            background: #e0e0e0;
+            color: #666;
+            border: none;
+            border-radius: 15px;
+            font-size: 0.85em;
+            cursor: pointer;
+            transition: background 0.2s, color 0.2s;
+        }
+
+        .region-filter-btn.active {
+            background: #27ae60;
+            color: white;
+        }
         
         #map {
             height: 400px;
@@ -199,7 +215,7 @@
 <body>
     <div class="container">
         <div class="header">
-            <h1>🌿 NDVI Time Series Analysis</h1>
+            <h1>🌿 NDVI Time Series Predicts</h1>
             <p>Phân tích chỉ số thực vật theo thời gian từ dữ liệu Sentinel-2</p>
         </div>
         
@@ -208,7 +224,7 @@
             <a href="/training" style="padding: 10px 20px; background: #f093fb; color: white; border-radius: 8px; text-decoration: none; font-weight: 600;">🎓 Training</a>
             <a href="/prediction" style="padding: 10px 20px; background: #4facfe; color: white; border-radius: 8px; text-decoration: none; font-weight: 600;">🗺️ Prediction</a>
             <a href="/batch" style="padding: 10px 20px; background: #764ba2; color: white; border-radius: 8px; text-decoration: none; font-weight: 600;">🚀 Batch Processing</a>
-            <a href="/ndvi" style="padding: 10px 20px; background: #2ecc71; color: white; border-radius: 8px; text-decoration: none; font-weight: 600;">🌿 NDVI Analysis (Active)</a>
+            <a href="/ndvi" style="padding: 10px 20px; background: #2ecc71; color: white; border-radius: 8px; text-decoration: none; font-weight: 600;">🌿 NDVI Predicts (Active)</a>
             <a href="/reports" style="padding: 10px 20px; background: #ff6b6b; color: white; border-radius: 8px; text-decoration: none; font-weight: 600;">📝 Reports</a>
         </div>
         
@@ -251,7 +267,7 @@
                     
                     <!-- Region Filters -->
                     <div id="regionFilterContainer" style="display: flex; gap: 5px; flex-wrap: wrap; margin-bottom: 10px;">
-                        <button type="button" class="region-filter-btn" data-region="all" style="padding: 3px 10px; background: #667eea; color: white; border: none; border-radius: 15px; font-size: 0.85em; cursor: pointer;">Tất cả</button>
+                        <button type="button" class="region-filter-btn active" data-region="all">Tất cả</button>
                     </div>
                     
                     <select id="provinceSelect" style="width: 100%; padding: 10px; border: 2px solid #e0e0e0; border-radius: 5px;">
@@ -397,7 +413,18 @@
         let allProvincesMerged = {};
         let currentProvinceName = '';
         let currentProvinceMode = '63';
+        let currentRegionFilter = 'all';
         let currentAnalysisMode = 'historical';
+
+        // Check if date range goes into the future
+        function isFutureRange(startDate, endDate) {
+            if (!startDate || !endDate) return false;
+            const today = new Date();
+            today.setHours(0, 0, 0, 0);
+            const start = new Date(startDate);
+            const end = new Date(endDate);
+            return start > today || end > today;
+        }
         
         // Switch between historical and forecast mode
         function switchMode(mode) {
@@ -453,6 +480,25 @@
         
         // Execute analysis based on current mode
         async function executeAnalysis() {
+            const startDate = document.getElementById('startDate').value;
+            const endDate = document.getElementById('endDate').value;
+
+            if (!startDate || !endDate) {
+                showError('Vui lòng chọn khoảng thời gian!');
+                return;
+            }
+
+            // If user picks a future range while in historical mode, switch to forecast automatically
+            if (currentAnalysisMode === 'historical' && isFutureRange(startDate, endDate)) {
+                switchMode('forecast');
+                const methodAlert = document.getElementById('methodAlert');
+                methodAlert.innerHTML = '<strong>🔄 Đã chuyển sang chế độ Dự đoán tương lai</strong><br>' +
+                    'Khoảng thời gian nằm trong tương lai, hệ thống tự dùng mô hình forecast để tránh lỗi "No Sentinel-2 data".';
+                document.getElementById('errorContainer').style.display = 'none';
+                await forecastNDVI();
+                return;
+            }
+
             if (currentAnalysisMode === 'historical') {
                 await predictNDVI();
             } else {
@@ -682,10 +728,11 @@
             
             // Update stats - handle both n_images (historical) and n_forecast_points (forecast)
             const nImages = data.n_images || data.n_forecast_points || 0;
+            const safeStat = (val) => (val === undefined || val === null || Number.isNaN(val) ? '-' : Number(val).toFixed(3));
             document.getElementById('statImages').textContent = nImages;
-            document.getElementById('statAvgNDVI').textContent = data.mean_ndvi.toFixed(3);
+            document.getElementById('statAvgNDVI').textContent = safeStat(data.mean_ndvi);
-            document.getElementById('statMinNDVI').textContent = data.min_ndvi.toFixed(3);
+            document.getElementById('statMinNDVI').textContent = safeStat(data.min_ndvi);
-            document.getElementById('statMaxNDVI').textContent = data.max_ndvi.toFixed(3);
+            document.getElementById('statMaxNDVI').textContent = safeStat(data.max_ndvi);
             
             // Create chart
             createChart(data);
@@ -699,6 +746,21 @@
             if (ndviChart) {
                 ndviChart.destroy();
             }
+
+            if (!data.timeseries || data.timeseries.length === 0) {
+                showError('Không có dữ liệu để vẽ biểu đồ.');
+                return;
+            }
+
+            const pickValue = (item, keys) => {
+                for (const key of keys) {
+                    const val = item[key];
+                    if (val !== undefined && val !== null && !Number.isNaN(val)) {
+                        return Number(val);
+                    }
+                }
+                return null;
+            };
             
             // Format dates with more detail (dd/MM/yyyy)
             const dates = data.timeseries.map(item => {
@@ -708,9 +770,9 @@
                 const year = d.getFullYear();
                 return `${day}/${month}/${year}`;
             });
-            const ndviValues = data.timeseries.map(item => item.mean_ndvi);
+            const ndviValues = data.timeseries.map(item => pickValue(item, ['mean_ndvi', 'ndvi_mean', 'ndvi']));
-            const minValues = data.timeseries.map(item => item.min_ndvi);
+            const minValues = data.timeseries.map(item => pickValue(item, ['min_ndvi', 'ndvi_min']));
-            const maxValues = data.timeseries.map(item => item.max_ndvi);
+            const maxValues = data.timeseries.map(item => pickValue(item, ['max_ndvi', 'ndvi_max']));
             
             ndviChart = new Chart(ctx, {
                 type: 'line',
@@ -837,7 +899,7 @@
                 allProvinces = await response63.json();
                 allProvincesMerged = await response32.json();
                 
-                populateProvinceSelect();
+                populateProvinceSelect(currentRegionFilter);
                 populateRegionButtons();
             } catch (error) {
                 console.error('Error loading provinces:', error);
@@ -951,10 +1013,19 @@
                 button.className = 'region-filter-btn';
                 button.dataset.region = region;
                 button.textContent = region;
-                button.style.cssText = 'padding: 3px 10px; background: #e0e0e0; color: #666; border: none; border-radius: 15px; font-size: 0.85em; cursor: pointer;';
+                if (region === currentRegionFilter) {
+                    button.classList.add('active');
+                }
                 container.appendChild(button);
             });
             
+            // Re-apply active state for "Tất cả"
+            if (currentRegionFilter === 'all') {
+                allButton.classList.add('active');
+            } else {
+                allButton.classList.remove('active');
+            }
+
             setupRegionFilters();
         }
         
@@ -964,16 +1035,10 @@
             
             filterButtons.forEach(btn => {
                 btn.addEventListener('click', function() {
-                    filterButtons.forEach(b => {
+                    filterButtons.forEach(b => b.classList.remove('active'));
-                        b.style.background = '#e0e0e0';
+                    this.classList.add('active');
-                        b.style.color = '#666';
+                    currentRegionFilter = this.dataset.region;
-                    });
+                    populateProvinceSelect(currentRegionFilter);
-                    
-                    this.style.background = '#27ae60';
-                    this.style.color = 'white';
-                    
-                    const region = this.dataset.region;
-                    populateProvinceSelect(region);
                 });
             });
         }