I was going to use a standard array, I need it to x number of rows and be 2 columns
double[,] members = new double[x, 2];
and loop through all the results I get back and then add them to the array, but as the number of results can change I don't want to pre-define the size of the array...
members[0, 0] = cost;
members[x, 1] = tax;
x++;
I was looking at resizing the array but would it be easier just using an arraylist or list of lists for this?
I would have a list of some type, where that type has a cost and a tax.
If space is a premium, it could be a struct with two doubles (btw, would decimal be more appropriate? It is almost always is a better choice for money), but if so: make it immutable.
For example:
public struct YourType {
private readonly decimal cost, tax;
public decimal Cost { get { return cost; } }
public decimal Tax { get { return tax; } }
public YourType(decimal cost, decimal tax) {
this.cost = cost;
this.tax = tax;
}
}
With:
List<YourType> list = new List<YourType>();
Perhaps the number of items in the second dimension is fixed, or at least, is predictable, so if you're using Microsoft .NET Framework 4.0, you can use a list of tuples:
List<Tuple<double, double>> list = new List<Tuple<int, int>>();
// In your loop...
list.Add(Tuple.Create(cost, tax));
// ... later when retrieving some tuple...
int item0cost = list[0].Item1;
int item0tax = list[0].Item2;
You can always implement some kind of value object class which has Cost and Tax properties, but using tuples you can do it as easy as I shown you in my code sample.
You should use a List<List<int>>.
Use a List<T>.
From msdn:
ArrayList has been replaced by
List<T>which is faster, and has more features.
T could be a List<double> or a double[2] or a struct or a 2-tuple... but as stated by Marc, using a List<List<double>> is over-doing things if the nested List is destined to always have two items.
Tuple<,>, but that isn't the only kind of tuple. Actually, an array is marginally more expensive due to storing the length. Oddly enough, thinking about it, you could just use a double[] and double the length, so item "3" is actually composed of arr[6] and arr[7].