Dynamic array of floats?

Dynamic array of floats?

• DOSrelic

  I would be tempted to say "use a generic list": List<float> provides both good performance and the ToArray() method that will return a float[] whenever you need to call your dll.

  Anyway, something in your post makes me wonder. What do you mean exactly by "resized at runtume based on some user-selectable parameter" If you means something like (C):

  float* myArray = (float*) malloc (numberOfElements * sizeof (float));

  then you can simply write (C#)

  float [] myArray = new float [numberOfElements];

  HTH
  --mc

  
  

• Peter Lillevold

  Hi LKeene,
  Yes, you can create bidimensional arrays like that... and yes, generic lists are a new feature of 2.0

  As for your last question, I'm not sure it's the best way, but you can get an array from the list with a single call to List<T>.ToArray(), and simply pass the array.

  HTH
  --mc

  
  

• FSchmid

  I'd like to avoid ArrayList because I think the overhead will be a problem (the app is a little performance sensitive - I've hand written some fast functions in C that I'd like to use in C#).

  I found the following article:http://msdn.microsoft.com/library/default.asp url=/library/en-us/dnnetcomp/html/netcfmarshallingtypes.asp

  Granted it deals with compact framework but I think the same will work in "regular" .net framework. From the article:

  "Array parameters have different representations in managed and unmanaged environments. In C/C++, arrays are representations as pointers to a contiguous region of memory with the array elements addressed as offsets from this pointer, starting with zero. The marshaler in the .NET Compact Framework Common Language Runtime (CLR) ensures that managed arrays adhere to this format when passed to unmanaged code. In this C/C++ example, an array is passed into a function and is searched to locate the smallest value:

  extern "C" _declspec(dllexport) int MinArray(int* pData, int length)
  {
    // Initialise minData to the first element of the pData Array
    int minData = pData[0];
    int pos;
  
    // Loop through the array
    for(pos = 1; pos < length; pos++)
    {
      // If the current element is less than minData, 
      // set minData to the value of current element
      if(pData[pos] < minData)
        minData = pData[pos];
    }
  
    return minData;
  }
  

  The following code declares the above method and calls it from C#:

  [DllImport("MarshalArray.dll")]
  extern static int MinArray(int[] pData, int length);
  .
  .
  .
  int[] sampleData = int[] {5, 1, 3 };
  int result = MinArray(sampleData, sampleData.Length);
  MessageBox.Show(String.Format("Smallest integer is {0}, result));

  
  

• IGiberson

  Could you also use an ArrayList, considering it stores an array of objects That would allow you to dynamically resize the list at any time.
  
  Or am I barking up the wrong tree
  
  

• Wayne R

  Thanks Mario, that is indeed what I was getting at. Can I assume the following is correct as well:

  float [,] myArray = new float[numberOfElements,numberOfElements];

  Generics are a new feature introduced in 2.0 aren't they What are it's P/Invoke data marshalling issues i.e. does it have an equivalent type in unmanaged code

  Thanks everyone.

  
  
• Dynamic array of floats?