Feature Vector<Type, NN> does not work as AutoML validation data

[thoron]

System information

• Windows 10.0.21364.0
• .NET SDK Version: 5.0.202

Issue

• What did you do?
  I want to use AutoML and having the feature column as a Vector<Type, NN> and add validation data to the experiment as in the exposed method:

public ExperimentResult<TMetrics> Execute(IDataView trainData, IDataView validationData, string labelColumnName = "Label", IEstimator<ITransformer> preFeaturizer = null, IProgress<RunDetail<TMetrics>> progressHandler = null)

Example code:

_mlContext.Auto()
    .CreateRegressionExperiment(experimentSettings)
    .Execute(_trainDataView, _testDataView, progressHandler: experimentProgress, labelColumnName: "LabelFeaturized");

• What happened?
  The experiment fails with to following exception:

System.ArgumentException: Training data and validation data schemas do not match. Column 'Features' is of type Vector<Single, 5> in train data, and type Vector<Single, 5> in validation data. (Parameter 'validationData')
    at Microsoft.ML.AutoML.UserInputValidationUtil.ValidateValidationData(IDataView trainData, IDataView validationData)
at Microsoft.ML.AutoML.UserInputValidationUtil.ValidateExperimentExecuteArgs(IDataView trainData, ColumnInformation columnInformation, IDataView validationData, TaskKind task)
at Microsoft.ML.AutoML.ExperimentBase`2.ExecuteTrainValidate(IDataView trainData, ColumnInformation columnInfo, IDataView validationData, IEstimator`1 preFeaturizer, IProgress`1 progressHandler)
at Microsoft.ML.AutoML.ExperimentBase`2.Execute(IDataView trainData, IDataView validationData, ColumnInformation columnInformation, IEstimator`1 preFeaturizer, IProgress`1 progressHandler)
at Microsoft.ML.AutoML.ExperimentBase`2.Execute(IDataView trainData, IDataView validationData, String labelColumnName, IEstimator`1 preFeaturizer, IProgress`1 progressHandler)

• What did you expect?
  The validation data to be used in the experiment and the Vector typed Feature column to be an acceptable datatype.

Perhaps this line needs to account for Vector<Type,NN> ?

machinelearning/src/Microsoft.ML.AutoML/Utils/UserInputValidationUtil.cs

Line 220 in 04dda55

if (trainCol.Type != validCol.Value.Type)

[michaelgsharp]

Yeah, good catch. I don't think that will be too hard to fix.