C# Splitting An Array

I also want to add a solution to split an array into several smaller arrays containing a determined number of cells.

A nice way would be to create a generic/extension method to split any array. This is mine:

/// <summary>
/// Splits an array into several smaller arrays.
/// </summary>
/// <typeparam name="T">The type of the array.</typeparam>
/// <param name="array">The array to split.</param>
/// <param name="size">The size of the smaller arrays.</param>
/// <returns>An array containing smaller arrays.</returns>
public static IEnumerable<IEnumerable<T>> Split<T>(this T[] array, int size)
{
    for (var i = 0; i < (float)array.Length / size; i++)
    {
        yield return array.Skip(i * size).Take(size);
    }
}

Moreover, this solution is deferred. Then, simply call split(size) on your array.

var array = new byte[] {10, 20, 30, 40, 50};
var splitArray = array.Split(2);

Have fun :)

Use a generic split method:

public static void Split<T>(T[] source, int index, out T[] first, out T[] last)
{
    int len2 = source.Length - index;
    first = new T[index];
    last = new T[len2];
    Array.Copy(source, 0, first, 0, index);
    Array.Copy(source, index, last, 0, len2);
}

I had an issue with Linq's Skip() and Take() functions when dealing with arrays with massive amounts of elements (i.e. byte arrays), where element counts are in the millions.

This approach dramatically reduced split execute times for me.

public static IEnumerable<IEnumerable<T>> Split<T>(this ICollection<T> self, int chunkSize)
{
    var splitList = new List<List<T>>();
    var chunkCount = (int)Math.Ceiling((double)self.Count / (double)chunkSize);

    for(int c = 0; c < chunkCount; c++)
    {
        var skip = c * chunkSize;
        var take = skip + chunkSize;
        var chunk = new List<T>(chunkSize);

        for(int e = skip; e < take && e < self.Count; e++)
        {
            chunk.Add(self.ElementAt(e));
        }

        splitList.Add(chunk);
    }

    return splitList;
}

It can be made in a form of a extension with usage of Range and Tuples

public static (T[], T[]) Split<T>(this T[] arr, int index) =>  (arr[..index], arr[index..]);

Example of usage:

var arr = new[] { 1, 2, 3, 4, 5, 6, 7, 8, 9, 0 };
var (arr1, arr2) = arr.Split(5);

If you don't have Linq, you can use Array.Copy:

public void Split(ref UList list)
{
    string[] s = list.mylist.ToArray();

    //split the array into top and bottom halfs
    string[] top = new string[s.Length / 2];
    string[] bottom = new string[s.Length - s.Length / 2];
    Array.Copy(s, top, top.Length);
    Array.Copy(s, top.Length, bottom, 0, bottom.Length);

    Console.WriteLine("Top: ");
    foreach (string item in top) Console.WriteLine(item);
    Console.WriteLine("Bottom: ");
    foreach (string item in bottom) Console.WriteLine(item);
}

Why are you passing the UList as ref? There doesn't appear to be a need for that.

I would use a generic Split method if I needed to do this:

public void Split<T>(T[] array, out T[] left, out T[] right)
{
    left = new T[array.Length / 2];
    right = new T[array.Length - left.Length];

    Array.Copy(array, left, left.Length);
    Array.Copy(array, left.Length, right, 0, right.Length);
}

I think what you're looking for is the Array class, specifically the Array.Copy static method. You can think of that class as containing the methods that would be instance methods of arrays if C# arrays had methods.

Why don't you allocate two arrays and copy the contents ?

EDIT: here you go:

        String[] origin = new String[4];
        origin[0] = "zero";
        origin[1] = "one";
        origin[2] = "two";
        origin[3] = "three";

        Int32 topSize = origin.Length / 2;
        Int32 bottomSize = origin.Length - topSize;
        String[] sTop = new String[topSize];
        String[] sBottom = new String[bottomSize];
        Array.Copy(origin, sTop, topSize);
        Array.Copy(origin, topSize , sBottom, 0, bottomSize);

If functional paradigm is a concern, this might help:

   public static IEnumerable<IEnumerable<T>> Split<T>(this IEnumerable<T> seq, Int32 sizeSplits) {
     Int32 numSplits = (seq.Count() / sizeSplits) + 1;
     foreach ( Int32 ns in Enumerable.Range(start: 1, count: numSplits) ) {
        (Int32 start, Int32 end) = GetIndexes(ns);
        yield return seq.Where((_, i) => (start <= i && i <= end));
     }

     (Int32 start, Int32 end) GetIndexes(Int32 numSplit) {
        Int32 indBase1 = numSplit * sizeSplits;
        Int32 start = indBase1 - sizeSplits;
        Int32 end = indBase1 - 1;
        return (start, end);
     }
  }