KeyToVectorMappingTransformer: Index was outside the bounds of the array.

[lisahua]

System information

• OS version/distro: 1.0.0
• .NET Version (eg., dotnet --info): 4.6.2

Issue

• What did you do? Apply Normalization superviseBin and OneHot,
• What happened? Got "Index was outside the bounds of the array."
• What did you expect? No error

Source code / logs

Please paste or attach the code or logs or traces that would be helpful to diagnose the issue you are reporting.

            MLContext mlContext = new MLContext(seed: 0);
            var dataPath = "test1.csv";
            var featureName = "Features";
            var loader = mlContext.Data.CreateTextLoader(new[] 
            {
                new TextLoader.Column("int1", DataKind.Int64, 0),
                new TextLoader.Column("int2", DataKind.Int64, 1),
                new TextLoader.Column("Label", DataKind.Boolean, 2),
            }, hasHeader: true, separatorChar: ',');

            var data = loader.Load(dataPath);
            var learningPipeline = mlContext.Transforms.Conversion.ConvertType("int1", outputKind: DataKind.Single)
                    .Append(mlContext.Transforms.Conversion.ConvertType("int2", outputKind: DataKind.Single))
                    .Append(mlContext.Transforms.Concatenate(featureName, new string[] { "int1", "int2" }))
                    .Append(mlContext.Transforms.NormalizeSupervisedBinning(featureName, fixZero: false, maximumBinCount: 5, labelColumnName: "Label"))
                    .Append(mlContext.Transforms.Categorical.OneHotEncoding(featureName, outputKind: OneHotEncodingEstimator.OutputKind.Indicator));
            learningPipeline.Fit(data).Transform(data).Preview();

int1, int2, label
301, 2000, true
450, 3000, true
-300, 4000, true
300, 2000, false
115, 2000, false
115, 2000, false

I think it is related to issue 1751 And based on the discussion for this issue, I tried

• adding MapKeyToValue() but no help.
• OneHotEncode cannot be easily removed, we want to treat binning as categorical feature.

            MLContext mlContext = new MLContext(seed: 0);
            var dataPath = "test1.csv";
            var featureName = "Features";
            var loader = mlContext.Data.CreateTextLoader(new[] 
            {
                new TextLoader.Column("int1", DataKind.Int64, 0),
                new TextLoader.Column("int2", DataKind.Int64, 1),
                new TextLoader.Column("Label", DataKind.Boolean, 2),
            }, hasHeader: true, separatorChar: ',');

            var data = loader.Load(dataPath);
            var learningPipeline = mlContext.Transforms.Conversion.ConvertType("int1", outputKind: DataKind.Single)
                    .Append(mlContext.Transforms.Conversion.ConvertType("int2", outputKind: DataKind.Single))
                    .Append(mlContext.Transforms.Conversion.MapValueToKey("Label"))
                    .Append(mlContext.Transforms.Concatenate(featureName, new string[] { "int1", "int2" }))
                    .Append(mlContext.Transforms.NormalizeSupervisedBinning(featureName, fixZero: false, maximumBinCount: 5, labelColumnName: "Label"))
                    .Append(mlContext.Transforms.Categorical.OneHotEncoding(featureName, outputKind: OneHotEncodingEstimator.OutputKind.Indicator))
                    .Append(mlContext.Transforms.Conversion.MapKeyToValue("Label"));
            learningPipeline.Fit(data).Transform(data).Preview();