// Licensed to the .NET Foundation under one or more agreements.
// The .NET Foundation licenses this file to you under the MIT license.
// See the LICENSE file in the project root for more information.
 
using System;
using System.Collections.Generic;
using System.Linq;
using Microsoft.ML;
using Microsoft.ML.CommandLine;
using Microsoft.ML.Data;
using Microsoft.ML.Internal.Internallearn;
using Microsoft.ML.Runtime;
using Microsoft.ML.Trainers.Ensemble;
 
[assembly: LoadableClass(typeof(RegressionEnsembleTrainer), typeof(RegressionEnsembleTrainer.Arguments),
    new[] { typeof(SignatureRegressorTrainer), typeof(SignatureTrainer) },
    RegressionEnsembleTrainer.UserNameValue,
    RegressionEnsembleTrainer.LoadNameValue)]
 
[assembly: LoadableClass(typeof(RegressionEnsembleTrainer), typeof(RegressionEnsembleTrainer.Arguments), typeof(SignatureModelCombiner),
    "Regression Ensemble Model Combiner", RegressionEnsembleTrainer.LoadNameValue)]
 
namespace Microsoft.ML.Trainers.Ensemble
{
    using TScalarPredictor = IPredictorProducing<Single>;
    using TScalarTrainer = ITrainerEstimator<ISingleFeaturePredictionTransformer<IPredictorProducing<float>>, IPredictorProducing<float>>;
 
    internal sealed class RegressionEnsembleTrainer : EnsembleTrainerBase<Single,
       IRegressionSubModelSelector, IRegressionOutputCombiner>,
       IModelCombiner
    {
        public const string LoadNameValue = "EnsembleRegression";
        public const string UserNameValue = "Regression Ensemble (bagging, stacking, etc)";
 
        public sealed class Arguments : ArgumentsBase
        {
            [Argument(ArgumentType.Multiple, HelpText = "Algorithm to prune the base learners for selective Ensemble", ShortName = "pt", SortOrder = 4)]
            [TGUI(Label = "Sub-Model Selector(pruning) Type", Description = "Algorithm to prune the base learners for selective Ensemble")]
            public ISupportRegressionSubModelSelectorFactory SubModelSelectorType = new AllSelectorFactory();
 
            [Argument(ArgumentType.Multiple, HelpText = "Output combiner", ShortName = "oc", SortOrder = 5)]
            [TGUI(Label = "Output combiner", Description = "Output combiner type")]
            public ISupportRegressionOutputCombinerFactory OutputCombiner = new MedianFactory();
 
            // REVIEW: If we make this public again it should be an *estimator* of this type of predictor, rather than the (deprecated) ITrainer.
            [Argument(ArgumentType.Multiple, HelpText = "Base predictor type", ShortName = "bp,basePredictorTypes", SortOrder = 1, Visibility = ArgumentAttribute.VisibilityType.CmdLineOnly, SignatureType = typeof(SignatureRegressorTrainer))]
            public IComponentFactory<TScalarTrainer>[] BasePredictors;
 
            internal override IComponentFactory<TScalarTrainer>[] GetPredictorFactories() => BasePredictors;
 
            public Arguments()
            {
                BasePredictors = new[]
                {
                    ComponentFactoryUtils.CreateFromFunction(env => new OnlineGradientDescentTrainer(env, LabelColumnName, FeatureColumnName))
                };
            }
        }
 
        private readonly ISupportRegressionOutputCombinerFactory _outputCombiner;
 
        public RegressionEnsembleTrainer(IHostEnvironment env, Arguments args)
            : base(args, env, LoadNameValue)
        {
            SubModelSelector = args.SubModelSelectorType.CreateComponent(Host);
            _outputCombiner = args.OutputCombiner;
            Combiner = args.OutputCombiner.CreateComponent(Host);
        }
 
        private RegressionEnsembleTrainer(IHostEnvironment env, Arguments args, PredictionKind predictionKind)
            : this(env, args)
        {
            Host.CheckParam(predictionKind == PredictionKind.Regression, nameof(PredictionKind));
        }
 
        private protected override PredictionKind PredictionKind => PredictionKind.Regression;
 
        private protected override IPredictor CreatePredictor(List<FeatureSubsetModel<float>> models)
        {
            return new EnsembleModelParameters(Host, PredictionKind, CreateModels<TScalarPredictor>(models), Combiner);
        }
 
        public IPredictor CombineModels(IEnumerable<IPredictor> models)
        {
            Host.CheckValue(models, nameof(models));
            Host.CheckParam(models.All(m => m is TScalarPredictor), nameof(models));
 
            var combiner = _outputCombiner.CreateComponent(Host);
            var p = models.First();
 
            var predictor = new EnsembleModelParameters(Host, p.PredictionKind,
                    models.Select(k => new FeatureSubsetModel<float>((TScalarPredictor)k)).ToArray(), combiner);
 
            return predictor;
        }
    }
}