Thread: Dynamic arrays

I have a piece of code which performs operations on a user defined amount of channels.
I am trying to use an array to handle the channels (following Salem's advice) which works well when testing for an array of fixed length as below:

Code:

class class_name: 
	.
	.

	protected:

		virtual void function(int n, float *const *in, float *const *out);
		float *Outchan[3];     // Declare Array for 4 Channels

	.
	.


void class_name::function(int n, float *const *in, float *const *out)
{
	int i,nofout;
	for (nofout=0; nofout<num_channels;nofout++)
	{OutChan[nofout] = out[nofout];}   // Set number of outputs in signal vector

	if (First==1)
	{
		ULStat = cbGetStatus (BoardNum, &Status, &CurCount, &CurIndex,AIFUNCTION);
		for (nofout=0; nofout<num_channels;nofout++)
				{OutChan[nofout] = 0;} // Give each channel a value in turn
		}
		
		else First=0;
	}

	else
	{
		while (n>0)
		{
				for (nofout=0; nofout<num_channels;nofout++)  //adapt to varying channels
				{
					for (i=DuplicateSample;i>0;i--) //number of duplicate same sample
					{
						*OutChan[nofout]++ = ADData[OutCount];
					}
					*OutChan[nofout]++ = (ADData[OutCount++];
					if (OutCount==Count) OutCount=0;
				}
				
				n=n-(DuplicateSample+1);
		}
	}
}

However, I can't figure out how to adapt this to work with a dynamic array.
I replaced

Code:

float *Outchan[3];

with

Code:

float *OutChan;  // in class declaration
. 
OutChan= new float[num_channels];  // in the function

But the compiler returns conversion errors.

I suspect my phobia & lack of understanding of pointers are letting me down again!

Could someone please explain how I would correctly use a dynamic array here?
Thanks

You are creating an array of float pointers, so OutChan should be a double pointer:

Code:

float **OutChan;  // in class declaration
. 
OutChan= new float*[num_channels];  // in the function

Of course, there's little reason to use this kind of dynamic array, it will cause a lot more grief for you and your class when you can use vector instead:

Code:

std::vector<float *> OutChan;  // in class declaration
. 
class_name::class_name(int num_channels) : OutChan(num_channels)  // in constructor initialization list
{
  // ... you could also resize it later with OutChan.resize(num_channels);
}

Daved your solution worked a treat.

The tip on vectors looks very interesting too, I'll head over to the main site and read up on them.

Thanks a lot