RandomForest not serializable

[mmehrle]

I just attempted to serialize my random forest like this:

Serializer.Save<RandomForest>(forest, fileName); // forest is the obj, fileName is the path

Throws a RT error that it is not marked as serializable. Then I tried this:

using (FileStream fs = new FileStream(fileName, FileMode.Create)) { new BinaryFormatter().Serialize(fs, forest); }

I checked the definition and apparently it is 'ParallelOptions' breaking this. Not sure how to fix this (yet).

[mmehrle]

Okay, I think I was able to fix it. See the code below:

`
namespace Accord.MachineLearning.DecisionTrees
{
using System;
using System.Collections.Generic;
using System.Linq;
using System.Text;
using System.Threading.Tasks;
using System.Data;
using System.Runtime.Serialization;
using System.Runtime.Serialization.Formatters.Binary;
using System.IO;
using Accord.Statistics.Filters;
using Accord.Math;
using AForge;
using Accord.Statistics;
using System.Threading;

/// <summary>
///   Random Forest.
/// </summary>
/// 
/// <remarks>
/// <para>
///   Represents a random forest of <see cref="DecisionTree"/>s. For 
///   sample usage and example of learning, please see the documentation
///   page for <see cref="RandomForestLearning"/>.</para>
/// </remarks>
/// 
/// <seealso cref="DecisionTree"/>
/// <seealso cref="RandomForestLearning"/>
/// 
[Serializable]
public class RandomForest : MulticlassClassifierBase, IParallel
{
    private DecisionTree[] trees;
    **[NonSerialized]
    private ParallelOptions parallelOptions;**


    /// <summary>
    ///   Gets the trees in the random forest.
    /// </summary>
    /// 
    public DecisionTree[] Trees
    {
        get { return trees; }
    }

    /// <summary>
    ///   Gets the number of classes that can be recognized
    ///   by this random forest.
    /// </summary>
    /// 
    [Obsolete("Please use NumberOfOutputs instead.")]
    public int Classes { get { return NumberOfOutputs; } }

    /// <summary>
    ///   Gets or sets the parallelization options for this algorithm.
    /// </summary>
    ///
    **public ParallelOptions ParallelOptions { get { return parallelOptions; } set { parallelOptions = value; } }**

    /// <summary>
    /// Gets or sets a cancellation token that can be used
    /// to cancel the algorithm while it is running.
    /// </summary>
    /// 
    public CancellationToken Token
    {
        get { return ParallelOptions.CancellationToken; }
        set { ParallelOptions.CancellationToken = value; }
    }

    /// <summary>
    ///   Creates a new random forest.
    /// </summary>
    /// 
    /// <param name="trees">The number of trees in the forest.</param>
    /// <param name="classes">The number of classes in the classification problem.</param>
    /// 
    public RandomForest(int trees, int classes)
    {
        this.trees = new DecisionTree[trees];
        this.NumberOfOutputs = classes;
        this.ParallelOptions = new ParallelOptions();
    }

    /// <summary>
    ///   Computes the decision output for a given input vector.
    /// </summary>
    /// 
    /// <param name="data">The input vector.</param>
    /// 
    /// <returns>The forest decision for the given vector.</returns>
    /// 
    [Obsolete("Please use Decide() instead.")]
    public int Compute(double[] data)
    {
        return Decide(data);
    }


    /// <summary>
    /// Computes a class-label decision for a given <paramref name="input" />.
    /// </summary>
    /// <param name="input">The input vector that should be classified into
    /// one of the <see cref="ITransform.NumberOfOutputs" /> possible classes.</param>
    /// <returns>A class-label that best described <paramref name="input" /> according
    /// to this classifier.</returns>
    public override int Decide(double[] input)
    {
        int[] responses = new int[NumberOfOutputs];
        Parallel.For(0, trees.Length, ParallelOptions, i =>
        {
            int j = trees[i].Decide(input);
            Interlocked.Increment(ref responses[j]);
        });

        return responses.ArgMax();
    }

   [OnDeserializing()]
    internal void OnDeserializingMethod(StreamingContext context)
    {
        this.ParallelOptions = new ParallelOptions();
    }
}

}
`

[cesarsouza]

Many thanks!

[Rbcody]

Dear C?sar, I just wanted to express my appreciation for all of the work you have done with Accord.net<http://Accord.net>. It has been a tremendous help in exploring data for so many of us! Thank you! Chip Cody (Robert B. Cody)

…

Sent from my iPad On Dec 28, 2016, at 5:44 AM, C?sar Souza <[email protected]<mailto:[email protected]>> wrote: Many thanks! - You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub<#331 (comment)>, or mute the thread<https://github.com/notifications/unsubscribe-auth/AXIcPgmsa7Ru6fmPUlsAsaUhdzdxCJgKks5rMj2mgaJpZM4Kd5cr>.

[cesarsouza]

Thanks a lot. Your appreciation indeed really means a lot to me (and I am sure to other contributors of the project as well!).

Regards,
Cesar

[cesarsouza]

Integrated in release 3.4.0.