Samples: exceptions / nits (dotnet#124)

Author: daholste

src/Samples/AutoTrainBinaryClassification.cs

@@ -15,12 +15,11 @@ public class AutoTrainBinaryClassification
         private static string TrainDataPath = $"{BaseDatasetsLocation}/wikipedia-detox-250-line-data.tsv";
         private static string TestDataPath = $"{BaseDatasetsLocation}/wikipedia-detox-250-line-test.tsv";
         private static string ModelPath = $"{BaseDatasetsLocation}/SentimentModel.zip";
-        private static string LabelColumnName = "Label";
+        private static string LabelColumnName = "Sentiment";
 
         public static void Run()
         {
-            //Create ML Context with seed for repeteable/deterministic results
-            MLContext mlContext = new MLContext(seed: 0);
+            MLContext mlContext = new MLContext();
 
             // STEP 1: Infer columns
             var columnInference = mlContext.Data.InferColumns(TrainDataPath, LabelColumnName, '\t');
@@ -32,16 +31,16 @@ public static void Run()
 
             // STEP 3: Auto featurize, auto train and auto hyperparameter tuning
             Console.WriteLine($"Invoking BinaryClassification.AutoFit");
-            var autoFitResults = mlContext.BinaryClassification.AutoFit(trainDataView, timeoutInSeconds: 60);
+            var autoFitResults = mlContext.BinaryClassification.AutoFit(trainDataView, LabelColumnName, timeoutInSeconds: 60);
 
             // STEP 4: Print metric from the best model
             var best = autoFitResults.Best();
-            Console.WriteLine($"Accuracy of best model from validation data {best.Metrics.Accuracy}");
+            Console.WriteLine($"Accuracy of best model from validation data: {best.Metrics.Accuracy}");
 
             // STEP 5: Evaluate test data
             IDataView testDataViewWithBestScore = best.Model.Transform(testDataView);
-            var testMetrics = mlContext.Regression.Evaluate(testDataViewWithBestScore, label: DefaultColumnNames.Label, DefaultColumnNames.Score);
-            Console.WriteLine($"Accuracy of best model from test data {best.Metrics.Accuracy}");
+            var testMetrics = mlContext.BinaryClassification.EvaluateNonCalibrated(testDataViewWithBestScore, label: LabelColumnName, DefaultColumnNames.Score);
+            Console.WriteLine($"Accuracy of best model from test data: {best.Metrics.Accuracy}");
 
             // STEP 6: Save the best model for later deployment and inferencing
             using (var fs = File.Create(ModelPath))

src/Samples/AutoTrainMulticlassClassification.cs

@@ -19,8 +19,7 @@ public class AutoTrainMulticlassClassification
 
         public static void Run()
         {
-            //Create ML Context with seed for repeteable/deterministic results
-            MLContext mlContext = new MLContext(seed: 0);
+            MLContext mlContext = new MLContext();
 
             // STEP 1: Infer columns
             var columnInference = mlContext.Data.InferColumns(TrainDataPath, LabelColumnName, '\t');
@@ -32,16 +31,16 @@ public static void Run()
 
             // STEP 3: Auto featurize, auto train and auto hyperparameter tuning
             Console.WriteLine($"Invoking MulticlassClassification.AutoFit");
-            var autoFitResults = mlContext.MulticlassClassification.AutoFit(trainDataView, timeoutInSeconds: 1);
+            var autoFitResults = mlContext.MulticlassClassification.AutoFit(trainDataView, timeoutInSeconds: 60);
 
             // STEP 4: Print metric from the best model
             var best = autoFitResults.Best();
-            Console.WriteLine($"AccuracyMacro of best model from validation data {best.Metrics.AccuracyMacro}");
+            Console.WriteLine($"AccuracyMacro of best model from validation data: {best.Metrics.AccuracyMacro}");
 
             // STEP 5: Evaluate test data
             IDataView testDataViewWithBestScore = best.Model.Transform(testDataView);
-            var testMetrics = mlContext.Regression.Evaluate(testDataViewWithBestScore, label: DefaultColumnNames.Label, DefaultColumnNames.Score);
-            Console.WriteLine($"AccuracyMacro of best model from test data {best.Metrics.AccuracyMacro}");
+            var testMetrics = mlContext.MulticlassClassification.Evaluate(testDataViewWithBestScore, label: DefaultColumnNames.Label, DefaultColumnNames.Score);
+            Console.WriteLine($"AccuracyMacro of best model from test data: {best.Metrics.AccuracyMacro}");
 
             // STEP 6: Save the best model for later deployment and inferencing
             using (var fs = File.Create(ModelPath))

src/Samples/AutoTrainRegression.cs

@@ -19,8 +19,7 @@ static class AutoTrainRegression
 
         public static void Run()
         {
-            //Create ML Context with seed for repeteable/deterministic results
-            MLContext mlContext = new MLContext(seed: 0);
+            MLContext mlContext = new MLContext();
 
             // STEP 1: Common data loading configuration
             var columnInference = mlContext.Data.InferColumns(TrainDataPath, LabelColumnName, ',');
@@ -32,16 +31,16 @@ public static void Run()
 
             // STEP 3: Auto featurize, auto train and auto hyperparameter tuning
             Console.WriteLine($"Invoking Regression.AutoFit");
-            var autoFitResults = mlContext.Regression.AutoFit(trainDataView, LabelColumnName, timeoutInSeconds:1);
+            var autoFitResults = mlContext.Regression.AutoFit(trainDataView, LabelColumnName, timeoutInSeconds: 60);
 
             // STEP 4: Compare and print actual value vs predicted value for top 5 rows from validation data
             var best = autoFitResults.Best();
-            Console.WriteLine($"RSquared of best model from validation data {best.Metrics.RSquared}");
+            Console.WriteLine($"RSquared of best model from validation data: {best.Metrics.RSquared}");
 
             // STEP 5: Evaluate test data
             IDataView testDataViewWithBestScore = best.Model.Transform(testDataView);
             var testMetrics = mlContext.Regression.Evaluate(testDataViewWithBestScore, label: DefaultColumnNames.Label, DefaultColumnNames.Score);
-            Console.WriteLine($"RSquared of best model from test data {best.Metrics.RSquared}");
+            Console.WriteLine($"RSquared of best model from test data: {best.Metrics.RSquared}");
 
             // STEP 6: Save the best model for later deployment and inferencing
             using (var fs = File.Create(ModelPath))

src/Samples/Data/iris-test.txt

@@ -1,4 +1,4 @@
-#Label	Sepal length	Sepal width	Petal length	Petal width
+Label	Sepal length	Sepal width	Petal length	Petal width
 0	5.1	3.5	1.4	0.2
 0	4.9	3.0	1.4	0.2
 0	4.7	3.2	1.3	0.2

src/Samples/Data/iris-train.txt

@@ -1,4 +1,4 @@
-#Label	Sepal length	Sepal width	Petal length	Petal width
+Label	Sepal length	Sepal width	Petal length	Petal width
 0	5.4	3.7	1.5	0.2
 0	4.8	3.4	1.6	0.2
 0	4.8	3.0	1.4	0.1