Convert double to float by cast or Convert.ToSingle()?

Question

In C# I can convert doubles to floats by a cast (float) or by Convert.ToSingle().

double x = 3.141592653589793238463;
float a = (float)x;
float b = Convert.ToSingle(x);

a and b become equal.

Are there any differences between both techniques? Which one should I prefer and why?

Emperor

Answer 1

From the .NET reference source:

public static float ToSingle(double value)
{
     return (float)value;
}

So, your answer is that they are exactly the same, under the hood.

Any preference between the two is strictly a personal style choice. Personally, I would always use the cast as it is shorter and and just seems more idiomatic to me.

Portillo

Since there does not seem to be any reason at all to use ToSingle it's not personal choice but objectively better to use a cast. – Katinakatine

The Convert class is meant to be the language-neutral way of converting between the different base types of the .NET Framework. Not all languages that run on .NET actually have a cast operator like C#. For example: in VB.NET you have have CType(), CDbl(), DirectCast() and implicit conversion, none of which have the exact same semantics as the cast operator in C#. – Dangerfield

Answer 2

While they are exactly the same when casting a double, there is a difference if the double has first been cast to an object.

object x = 1.0;
float a = (float)x;              //InvalidCastException
float b = Convert.ToSingle(x);   //OK

The .NET reference source for how it is done is a few lines above the answer provided by @Glorin.

Crapshooter

Answer 3

Depending on the version of C#, the Convert methods test whether values ​​are null before converting, and if so, return 0 for primitive numbers. For the same case in an explicit conversion an exception is raised.

An alternative to using a Convert method is to use the ?? operator

float floatValue = nullableFloat ?? 0.0f;

Archaeo

Answer 4

I converted a double to float for a homework assignment using the following: { float squareRoot= (float)Math.Sqrt(input); return squareRoot; }

Compute