feat: models caching

.gitignore

@@ -3,6 +3,7 @@
 ##
 ## Get latest from `dotnet new gitignore`
 
+.models_cache
 .zed
 *.png
 cache

GSS2.Core/Analysis/AmineContent/AnalyzerService.cs

@@ -141,6 +141,35 @@ public partial class AnalyzerService
         var sampleRecords = _context.SampleRecords
             .Where(r => r.BrandId == brand.Id);
 
+        var dataSb = new System.Text.StringBuilder();
+        dataSb.AppendLine(brand.Id.ToString());
+        dataSb.AppendLine(brand.EditDateTime.ToString());
+        foreach (var record in sampleRecords.OrderBy(r => r.Id))
+        {
+            dataSb.AppendLine(record.Id.ToString());
+            dataSb.AppendLine(record.EditDateTime.ToString());
+        }
+        dataSb.AppendLine(separator.Id.ToString());
+        dataSb.AppendLine(separator.EditDateTime.ToString());
+        var dataHashCode = Convert.ToHexString(
+            System.Security.Cryptography.MD5.HashData(
+                System.Text.Encoding.UTF8.GetBytes(dataSb.ToString())
+            )
+        );
+
+        _logger.LogInformation("Хэш калибровочных записей: {}", dataHashCode);
+
+        Directory.CreateDirectory(".models_cache");
+        var cachedModelPath = Path.Combine(".models_cache", dataHashCode) + ".zip";
+
+        if (File.Exists(cachedModelPath))
+        {
+            _logger.LogInformation("Модель найдена к кэше");
+            var model = _ml.Model.Load(cachedModelPath, out _);
+            _clusterizationEngine = _ml.Model.CreatePredictionEngine<ClusterizationData, ClusterizationPrediction>(model);
+        }
+        else
+        {
             _logger.LogInformation("Найдено калибровочных записей проб для марки: {}", sampleRecords.Count());
             if (sampleRecords.Count() == 0)
                 throw new InvalidDataException($"Не найдено калибровочных записей для марки: \"{brand.BrandName} - {brand.MixtureName}\"");
@@ -202,9 +231,13 @@ public partial class AnalyzerService
             var model = pipeline.Fit(trainData);
             cancellationToken.ThrowIfCancellationRequested();
 
+            _ml.Model.Save(model, trainData.Schema, cachedModelPath);
+
             _clusterizationEngine = _ml.Model.CreatePredictionEngine<ClusterizationData, ClusterizationPrediction>(model);
+
             model.Dispose();
         }
+    }
     private async Task InitializeMlRegressionEngine(BrandRecord brand, CancellationToken cancellationToken)
     {
         _logger.LogInformation("Инициализация модели регрессии для марки \"{} - {}\"", brand.BrandName, brand.MixtureName);
@@ -217,6 +250,33 @@ public partial class AnalyzerService
         var sampleRecords = _context.SampleRecords
             .Where(r => r.BrandId == brand.Id);
 
+        var dataSb = new System.Text.StringBuilder();
+        dataSb.AppendLine(brand.Id.ToString());
+        dataSb.AppendLine(brand.EditDateTime.ToString());
+        foreach (var record in sampleRecords.OrderBy(r => r.Id))
+        {
+            dataSb.AppendLine(record.Id.ToString());
+            dataSb.AppendLine(record.EditDateTime.ToString());
+        }
+        var dataHashCode = Convert.ToHexString(
+            System.Security.Cryptography.MD5.HashData(
+                System.Text.Encoding.UTF8.GetBytes(dataSb.ToString())
+            )
+        );
+
+        _logger.LogInformation("Хэш калибровочных записей: {}", dataHashCode);
+
+        Directory.CreateDirectory(".models_cache");
+        var cachedModelPath = Path.Combine(".models_cache", dataHashCode) + ".zip";
+
+        if (File.Exists(cachedModelPath))
+        {
+            _logger.LogInformation("Модель найдена к кэше");
+            var model = _ml.Model.Load(cachedModelPath, out _);
+            _regressionEngine = _ml.Model.CreatePredictionEngine<RegressionData, RegressionPrediction>(model);
+        }
+        else
+        {
             _logger.LogInformation("Найдено калибровочных записей проб для марки: {}", sampleRecords.Count());
             if (sampleRecords.Count() == 0)
                 throw new InvalidDataException($"Не найдено калибровочных записей для марки: \"{brand.BrandName} - {brand.MixtureName}\"");
@@ -250,6 +310,7 @@ public partial class AnalyzerService
                 .ToList();
 
             _logger.LogInformation("Загружено векторов для обучения: {}", trainDataSet.Count());
+
             if (trainDataSet.Count() == 0)
                 throw new InvalidDataException($"Не найдено калибровочных данных для марки: \"{brand.BrandName} - {brand.MixtureName}\"");
             cancellationToken.ThrowIfCancellationRequested();
@@ -287,10 +348,13 @@ public partial class AnalyzerService
 
             _logger.LogInformation("Обучение модели");
             var model = pipeline.Fit(trainData);
+            _ml.Model.Save(model, trainData.Schema, cachedModelPath);
+
             _regressionEngine = _ml.Model.CreatePredictionEngine<RegressionData, RegressionPrediction>(model);
 
             model.Dispose();
         }
+    }
 
     public async Task<(Mat avgImage, Rect roi, Mat mask, IEnumerable<Point[]> contours, Mat result)?> Analyze(IEnumerable<ImageData> imagesData)
     {