migrate F# samples to 0.10 (dotnet#244)

* migrate F# samples to 0.10

* post merge updates

Author: kevmal

samples/fsharp/common_v0.9/ConsoleHelper.fs

@@ -4,10 +4,10 @@ module ConsoleHelper =
     open System
     open Microsoft.ML
     open Microsoft.ML.Data
-    open Microsoft.ML.Data
     open Microsoft.ML.Core.Data
     //open Microsoft.ML.Api
     open System.Reflection
+    open Microsoft.Data.DataView
 
     let printPrediction prediction =
         printfn "*************************************************"
@@ -139,7 +139,7 @@ module ConsoleHelper =
         // 'transformedData' is a 'promise' of data, lazy-loading. Let's actually read it.
         // Convert to an enumerable of user-defined type.
         let someRows = 
-            transformedData.AsEnumerable<'TObservation>(mlContext, reuseRowObject = false)
+            mlContext.CreateEnumerable<'TObservation>(transformedData, reuseRowObject = false)
             // Take the specified number of rows
             |> Seq.take numberOfRows
             // Convert to List

samples/fsharp/common_v0.9/ModelBuilder.fs

@@ -7,6 +7,7 @@ open Microsoft.ML.Transforms
 open Microsoft.ML.Data
 
 module ModelBuilder =
+    open Microsoft.Data.DataView
 
     let create (mlContext : MLContext) (pipeline : IEstimator<ITransformer>) =
         (mlContext, pipeline)

samples/fsharp/getting-started/CreditCardFraudDetection/CreditCardFraudDetection.sln renamed to samples/fsharp/getting-started/BinaryClassification_CreditCardFraudDetection/CreditCardFraudDetection.sln

@@ -10,7 +10,7 @@
   </ItemGroup>
 
   <ItemGroup>
-    <PackageReference Include="Microsoft.ML" Version="0.9.0" />
+    <PackageReference Include="Microsoft.ML" Version="0.10.0" />
   </ItemGroup>
 
 </Project>

samples/fsharp/getting-started/CreditCardFraudDetection/CreditCardFraudDetection/CreditCardFraudDetection.fsproj renamed to samples/fsharp/getting-started/BinaryClassification_CreditCardFraudDetection/CreditCardFraudDetection/CreditCardFraudDetection.fsproj

@@ -123,11 +123,11 @@ let main _ =
     let loaderArgs = TextLoader.Arguments()
     loaderArgs.Column <- columns
     loaderArgs.HasHeader <- true
-    loaderArgs.Separator <- ","
+    loaderArgs.Separators <- [| ',' |]
 
     let reader = TextLoader (mlContext, loaderArgs)
 
-    let classification = BinaryClassificationContext mlContext
+    let classification = BinaryClassificationCatalog mlContext
 
     (*
     Split the data 80:20 into train and test files, 
@@ -207,15 +207,15 @@ let main _ =
                 trainFile,
                 columnsPlus,                                                           
                 loaderArgs.HasHeader,
-                loaderArgs.Separator.ToCharArray().[0]
+                loaderArgs.Separators.[0]
                 )
 
         let testData = 
             mlContext.Data.ReadFromTextFile(
                 testFile,
                 columnsPlus,
                 loaderArgs.HasHeader,
-                loaderArgs.Separator.ToCharArray().[0]
+                loaderArgs.Separators.[0]
                 )
 
         trainData, testData
@@ -237,8 +237,8 @@ let main _ =
         |> fun x -> 
             x.Append (
                 mlContext.Transforms.Normalize (
-                    "Features", 
                     "FeaturesNormalizedByMeanVar", 
+                    "Features", 
                     NormalizingEstimator.NormalizerMode.MeanVariance
                     )
                 )
@@ -279,7 +279,7 @@ let main _ =
     let testData = mlContext.Data.ReadFromTextFile (testFile, columnsPlus, hasHeader = true, separatorChar = ',')
 
     printfn "Making predictions"
-    testData.AsEnumerable<TransactionObservation>(mlContext, reuseRowObject = false)
+    mlContext.CreateEnumerable<TransactionObservation>(testData, reuseRowObject = false)
     |> Seq.filter (fun x -> x.Label = true)
     // use 5 observations from the test data
     |> Seq.take 5

samples/fsharp/getting-started/CreditCardFraudDetection/CreditCardFraudDetection/Program.fs renamed to samples/fsharp/getting-started/BinaryClassification_CreditCardFraudDetection/CreditCardFraudDetection/Program.fs

@@ -2,7 +2,7 @@
 
 | ML.NET version | API type          | Status                        | App Type    | Data type | Scenario            | ML Task                   | Algorithms                  |
 |----------------|-------------------|-------------------------------|-------------|-----------|---------------------|---------------------------|-----------------------------|
-| v0.9           | Dynamic API | README.md updated | Console app | .tsv files | Sentiment Analysis | Two-class  classification | Linear Classification |
+| v0.10           | Dynamic API | README.md updated | Console app | .tsv files | Sentiment Analysis | Two-class  classification | Linear Classification |
 
 ------------------------------------
 
@@ -45,7 +45,7 @@ The initial code is similar to the following:
 ```fsharp
     // STEP 1: Common data loading configuration
     let textLoader =
-        mlContext.Data.CreateTextReader (
+        mlContext.Data.CreateTextLoader(
             columns = 
                 [|
                     TextLoader.Column("Label", Nullable DataKind.Bool, 0)
@@ -60,7 +60,7 @@ The initial code is similar to the following:
 
 
     // STEP 2: Common data process configuration with pipeline data transformations          
-    let dataProcessPipeline = mlContext.Transforms.Text.FeaturizeText("Text", "Features")
+    let dataProcessPipeline = mlContext.Transforms.Text.FeaturizeText("Features", "Text")
 
     // STEP 3: Set the training algorithm, then create and config the modelBuilder
     let trainer = mlContext.BinaryClassification.Trainers.FastTree(labelColumn = "Label", featureColumn = "Features")

samples/fsharp/getting-started/CreditCardFraudDetection/Data/creditcardfraud-dataset.zip renamed to samples/fsharp/getting-started/BinaryClassification_CreditCardFraudDetection/Data/creditcardfraud-dataset.zip

@@ -27,7 +27,7 @@ let buildTrainEvaluateAndSaveModel (mlContext : MLContext) =
     let testDataView = mlContext.Data.ReadFromTextFile<SentimentIssue>(testDataPath, hasHeader = true)
 
     // STEP 2: Common data process configuration with pipeline data transformations          
-    let dataProcessPipeline = mlContext.Transforms.Text.FeaturizeText("Text", "Features")
+    let dataProcessPipeline = mlContext.Transforms.Text.FeaturizeText("Features", "Text")
 
     // (OPTIONAL) Peek data (such as 2 records) in training DataView after applying the ProcessPipeline's transformations into "Features" 
     Common.ConsoleHelper.peekDataViewInConsole<SentimentIssue> mlContext trainingDataView dataProcessPipeline 2 |> ignore

samples/fsharp/getting-started/BinaryClassification_SentimentAnalysis/README.md

@@ -18,7 +18,7 @@
   </ItemGroup>
 
   <ItemGroup>
-    <PackageReference Include="Microsoft.ML" Version="$(MicrosoftMLVersion)" />
+    <PackageReference Include="Microsoft.ML" Version="0.10.0" />
   </ItemGroup>
 
   <ItemGroup>

samples/fsharp/getting-started/BinaryClassification_SentimentAnalysis/SentimentAnalysis/SentimentAnalysisConsoleApp/Program.fs

@@ -47,7 +47,7 @@ let main _argv =
     // Set up the MLContext, which is a catalog of components in ML.NET.
     let mlContext = MLContext(seed = Nullable 1)
     let reader = 
-        mlContext.Data.CreateTextReader(
+        mlContext.Data.CreateTextLoader(
             columns = 
                 [|
                     TextLoader.Column("LabelText" , Nullable DataKind.Text, 0)
@@ -58,12 +58,11 @@ let main _argv =
 
     let data = reader.Read(trainDataPath)
 
-    // Create the estimator which converts the text label to a key sorted by value (with ham < spam we get ham -> false and spam -> true)
-    // then featurizes the text, and add a linear trainer.
+    // Create the estimator which converts the text label to a bool then featurizes the text, and add a linear trainer.
     let estimator = 
         EstimatorChain()
-            .Append(mlContext.Transforms.Conversion.MapValueToKey("LabelText", "Label", sort = ValueToKeyMappingTransformer.SortOrder.Value))
-            .Append(mlContext.Transforms.Text.FeaturizeText("Message","Features"))
+            .Append(mlContext.Transforms.Conversion.ValueMap(["ham"; "spam"], [false; true],[| struct ("Label", "LabelText") |]))
+            .Append(mlContext.Transforms.Text.FeaturizeText("Features", "Message"))
             .AppendCacheCheckpoint(mlContext)
             .Append(mlContext.BinaryClassification.Trainers.StochasticDualCoordinateAscent("Label", "Features"))