stop shuffle rows in ITrainValidationDatasetManager (dotnet#6742)

LittleLittleCloud · web-flow · commit 8858ab646630 · 2023-06-23T02:25:08.000-07:00
* stop shuffle rows

* make file read only
diff --git a/src/Microsoft.ML.AutoML/AutoMLExperiment/IDatasetManager.cs b/src/Microsoft.ML.AutoML/AutoMLExperiment/IDatasetManager.cs
@@ -46,7 +46,7 @@ public interface ITrainValidateDatasetManager : IDatasetManager
     internal class TrainValidateDatasetManager : IDatasetManager, ITrainValidateDatasetManager
     {
         private ulong _rowCount;
-        private IDataView _trainDataset;
+        private readonly IDataView _trainDataset;
         private readonly IDataView _validateDataset;
         private readonly string _subSamplingKey = "TrainValidateDatasetSubsamplingKey";
         private bool _isInitialized = false;
@@ -94,7 +94,6 @@ public IDataView LoadValidateDataset(MLContext context, TrialSettings? settings)
         private void InitializeTrainDataset(MLContext context)
         {
             _rowCount = DatasetDimensionsUtil.CountRows(_trainDataset, ulong.MaxValue);
-            _trainDataset = context.Data.ShuffleRows(_trainDataset);
         }
     }
 

Original file line number	Diff line number	Diff line change
`@@ -46,7 +46,7 @@ public interface ITrainValidateDatasetManager : IDatasetManager`
`46`	`46`	`internal class TrainValidateDatasetManager : IDatasetManager, ITrainValidateDatasetManager`
`47`	`47`	`{`
`48`	`48`	`private ulong _rowCount;`
`49`		`- private IDataView _trainDataset;`
	`49`	`+ private readonly IDataView _trainDataset;`
`50`	`50`	`private readonly IDataView _validateDataset;`
`51`	`51`	`private readonly string _subSamplingKey = "TrainValidateDatasetSubsamplingKey";`
`52`	`52`	`private bool _isInitialized = false;`
`@@ -94,7 +94,6 @@ public IDataView LoadValidateDataset(MLContext context, TrialSettings? settings)`
`94`	`94`	`private void InitializeTrainDataset(MLContext context)`
`95`	`95`	`{`
`96`	`96`	`_rowCount = DatasetDimensionsUtil.CountRows(_trainDataset, ulong.MaxValue);`
`97`		`- _trainDataset = context.Data.ShuffleRows(_trainDataset);`
`98`	`97`	`}`
`99`	`98`	`}`
`100`	`99`