Concatenating two arrays results in two modified arrays instead of one C#

Question

I have a very strange Behavior when merging two arrays together:

Assumptions

I have a class Tensor which contains an array float[] and a function AddTensorElements:

class Tensor
{
    public float[] MovingAverage3h { get; set; }
    public float[] MovingAverage6h { get; set; }
    public float[] MovingAverage1d { get; set; }
    public void AddTensorElements(Tensor input)
    {
        if (this.MovingAverage3h == null)
        this.MovingAverage3h = input.MovingAverage3h;
        this.MovingAverage6h = input.MovingAverage6h;
        this.MovingAverage1d = input.MovingAverage1d;
        }
        else
       {
       this.MovingAverage3h = Concat(this.MovingAverage3h, input.MovingAverage6h);
       this.MovingAverage6h = Concat(this.MovingAverage6h, input.MovingAverage6h);
       this.MovingAverage1d = Concat(this.MovingAverage1d, input.MovingAverage1d);
       }
    private float[] Concat (float[] first, float[] second)
    {
        List<float> concatenated = new List<float>();
        concatenated.AddRange(first);
        concatenated.AddRange(second);
        //foreach (float value in first) concatenated.Add(value);
        //foreach (float value in second) concatenated.Add(value);
        float[] returnArray = concatenated.ToArray();
        return returnArray;
    }
}

Within my main program, I repeatedly add the tensor M6t to the base tensor Minutes30[i]

class TensorCreator
{

    private static List<Elements.Tensor> Minutes30 = new List<Elements.Tensor>();

    private static void AddValues(Tensor M6t)
    {
        // Fill Minutes 30
        Minutes30.Add(M6t);
        for (int i = CounterM30; i < Minutes30.Count-1; i += M6)
        {
            { } // Issue come up right here
            Minutes30[i].AddTensorElements(M6t);
            { } // Issue come up right here
        }
    }
    public static void AppendDataToTensor(Elements.Tensor queueElement)
    {
        // ...
        AddValues(M6Avg);
    }
}

Expected behavior vs actual behavior

The array within Minutes[i] expands
The array within M6t staysfloat[1]

So far so good, this works in a tiny separate test application

Within my actual application, the same code lets the baseTensor expand but also the input tensor gets expanded!

for (int i = CounterM30; i < Minutes30.Count-1; i += M6)
{
    // M6T.Length == 1;
    Minutes30[i].AddTensorElements(M6t);
    // M6T.Length == Minutes30[i].Length;
}

strangely, whitin AddtensorToElements() I can see the values changing as well:

Answer 1

The Issue lies right here:

Minutes30.Add(M6t);

This adds a reference of Class Tensor M6t to Minutes 30. The result is that Minutes30[i] gets concatenated with it self.

Solution: In class Tensor, add a Clone() method

public Tensor Clone()
{
    Tensor tensor = new Tensor();
    tensor.MovingAverage3h = this.MovingAverage3h.ToArray();
    tensor.MovingAverage6h = this.MovingAverage6h.ToArray();
    tensor.MovingAverage1d = this.MovingAverage1d.ToArray();
    return tensor;
}

then change

    Minutes30.Add(M6t);

to

    Minutes30.Add(M6t.Clone());