C# - how to determine whether a Type is a number

Question

Is there a way to determine whether or not a given .Net Type is a number? For example: System.UInt32/UInt16/Double are all numbers. I want to avoid a long switc

Answer 1

With C# 7 this method gives to me better performance than switch case on TypeCode and HashSet<Type>:

public static bool IsNumeric(this object o) => o is byte || o is sbyte || o is ushort || o is uint || o is ulong || o is short || o is int || o is long || o is float || o is double || o is decimal;

Tests are following:

public static class Extensions
{
    public static HashSet<Type> NumericTypes = new HashSet<Type>()
    {
        typeof(byte), typeof(sbyte), typeof(ushort), typeof(uint), typeof(ulong), typeof(short), typeof(int), typeof(long), typeof(decimal), typeof(double), typeof(float)
    };

    public static bool IsNumeric1(this object o) => NumericTypes.Contains(o.GetType());

    public static bool IsNumeric2(this object o) => o is byte || o is sbyte || o is ushort || o is uint || o is ulong || o is short || o is int || o is long || o is decimal || o is double || o is float;

    public static bool IsNumeric3(this object o)
    {
        switch (o)
        {
            case Byte b:
            case SByte sb:
            case UInt16 u16:
            case UInt32 u32:
            case UInt64 u64:
            case Int16 i16:
            case Int32 i32:
            case Int64 i64:
            case Decimal m:
            case Double d:
            case Single f:
                return true;
            default:
                return false;
        }
    }

    public static bool IsNumeric4(this object o)
    {
        switch (Type.GetTypeCode(o.GetType()))
        {
            case TypeCode.Byte:
            case TypeCode.SByte:
            case TypeCode.UInt16:
            case TypeCode.UInt32:
            case TypeCode.UInt64:
            case TypeCode.Int16:
            case TypeCode.Int32:
            case TypeCode.Int64:
            case TypeCode.Decimal:
            case TypeCode.Double:
            case TypeCode.Single:
                return true;
            default:
                return false;
        }
    }
}

class Program
{
    static void Main(string[] args)
    {           
        var count = 100000000;

        //warm up calls
        for (var i = 0; i < count; i++)
        {
            i.IsNumeric1();
        }
        for (var i = 0; i < count; i++)
        {
            i.IsNumeric2();
        }
        for (var i = 0; i < count; i++)
        {
            i.IsNumeric3();
        }
        for (var i = 0; i < count; i++)
        {
            i.IsNumeric4();
        }

        //Tests begin here
        var sw = new Stopwatch();
        sw.Restart();
        for (var i = 0; i < count; i++)
        {
            i.IsNumeric1();
        }
        sw.Stop();

        Debug.WriteLine(sw.ElapsedMilliseconds);

        sw.Restart();
        for (var i = 0; i < count; i++)
        {
            i.IsNumeric2();
        }
        sw.Stop();

        Debug.WriteLine(sw.ElapsedMilliseconds);

        sw.Restart();
        for (var i = 0; i < count; i++)
        {
            i.IsNumeric3();
        }
        sw.Stop();

        Debug.WriteLine(sw.ElapsedMilliseconds);

        sw.Restart();
        for (var i = 0; i < count; i++)
        {
            i.IsNumeric4();
        }
        sw.Stop();

        Debug.WriteLine(sw.ElapsedMilliseconds);
    }

Answer 2

Short answer: No.

Longer Answer: Nope.

The fact is that many different types in C# can contain numeric data. Unless you know what to expect (Int, Double, etc) you need to use the "long" case statement.

Answer 3

Modified skeet's and arviman's solution utilizing Generics, Reflection, and C# v6.0.

private static readonly HashSet<Type> m_numTypes = new HashSet<Type>
{
    typeof(int),  typeof(double),  typeof(decimal),
    typeof(long), typeof(short),   typeof(sbyte),
    typeof(byte), typeof(ulong),   typeof(ushort),
    typeof(uint), typeof(float),   typeof(BigInteger)
};

Followed By:

public static bool IsNumeric<T>( this T myType )
{
    var IsNumeric = false;

    if( myType != null )
    {
        IsNumeric = m_numTypes.Contains( myType.GetType() );
    }

    return IsNumeric;
}

Usage for (T item):

if ( item.IsNumeric() ) {}

null returns false.

Answer 4

oops! Misread the question! Personally, would roll with Skeet's.

---

hrm, sounds like you want to DoSomething on Type of your data. What you could do is the following

public class MyClass
{
    private readonly Dictionary<Type, Func<SomeResult, object>> _map = 
        new Dictionary<Type, Func<SomeResult, object>> ();

    public MyClass ()
    {
        _map.Add (typeof (int), o => return SomeTypeSafeMethod ((int)(o)));
    }

    public SomeResult DoSomething<T>(T numericValue)
    {
        Type valueType = typeof (T);
        if (!_map.Contains (valueType))
        {
            throw new NotSupportedException (
                string.Format (
                "Does not support Type [{0}].", valueType.Name));
        }
        SomeResult result = _map[valueType] (numericValue);
        return result;
    }
}

Answer 5

This may work as well. However, you may want to follow it up with a Type.Parse to cast it the way you want it afterwards.

public bool IsNumeric(object value)
{
    float testValue;
    return float.TryParse(value.ToString(), out testValue);
}

Answer 6

public static bool IsNumericType(Type type)
{
  switch (Type.GetTypeCode(type))
  {
    case TypeCode.Byte:
    case TypeCode.SByte:
    case TypeCode.UInt16:
    case TypeCode.UInt32:
    case TypeCode.UInt64:
    case TypeCode.Int16:
    case TypeCode.Int32:
    case TypeCode.Int64:
    case TypeCode.Decimal:
    case TypeCode.Double:
    case TypeCode.Single:
      return true;
    default:
      return false;
  }
}

Note about optimization removed (see enzi comments) And if you really want to optimize it (losing readability and some safety...):

public static bool IsNumericType(Type type)
{
  TypeCode typeCode = Type.GetTypeCode(type);
  //The TypeCode of numerical types are between SByte (5) and Decimal (15).
  return (int)typeCode >= 5 && (int)typeCode <= 15;
}