private protected CrossValidationResult[] CrossValidateTrain(IDataView data, IEstimator <ITransformer> estimator,
int numFolds, string samplingKeyColumn, int?seed = null)
{
Environment.CheckValue(data, nameof(data));
Environment.CheckValue(estimator, nameof(estimator));
Environment.CheckParam(numFolds > 1, nameof(numFolds), "Must be more than 1");
Environment.CheckValueOrNull(samplingKeyColumn);
DataOperationsCatalog.EnsureGroupPreservationColumn(Environment, ref data, ref samplingKeyColumn, seed);
var result = new CrossValidationResult[numFolds];
int fold = 0;
// Sequential per-fold training.
// REVIEW: we could have a parallel implementation here. We would need to
// spawn off a separate host per fold in that case.
foreach (var split in DataOperationsCatalog.CrossValidationSplit(Environment, data, numFolds, samplingKeyColumn))
{
var model = estimator.Fit(split.TrainSet);
var scoredTest = model.Transform(split.TestSet);
result[fold] = new CrossValidationResult(model, scoredTest, fold);
fold++;
}
return(result);
}
private protected CrossValidationResult[] CrossValidateTrain(IDataView data, IEstimator <ITransformer> estimator,
int numFolds, string samplingKeyColumn, int?seed = null)
{
Environment.CheckValue(data, nameof(data));
Environment.CheckValue(estimator, nameof(estimator));
Environment.CheckParam(numFolds > 1, nameof(numFolds), "Must be more than 1");
Environment.CheckValueOrNull(samplingKeyColumn);
DataOperationsCatalog.EnsureGroupPreservationColumn(Environment, ref data, ref samplingKeyColumn, seed);
Func <int, CrossValidationResult> foldFunction =
fold =>
{
var trainFilter = new RangeFilter(Environment, new RangeFilter.Options
{
Column = samplingKeyColumn,
Min = (double)fold / numFolds,
Max = (double)(fold + 1) / numFolds,
Complement = true
}, data);
var testFilter = new RangeFilter(Environment, new RangeFilter.Options
{
Column = samplingKeyColumn,
Min = (double)fold / numFolds,
Max = (double)(fold + 1) / numFolds,
Complement = false
}, data);
var model = estimator.Fit(trainFilter);
var scoredTest = model.Transform(testFilter);
return(new CrossValidationResult(model, scoredTest, fold));
};
// Sequential per-fold training.
// REVIEW: we could have a parallel implementation here. We would need to
// spawn off a separate host per fold in that case.
var result = new CrossValidationResult[numFolds];
for (int fold = 0; fold < numFolds; fold++)
{
result[fold] = foldFunction(fold);
}
return(result);
}
