What is the best practice to make division return double in C#

Question

In c# when you want to divide the result of a method such as below, what is the best way to force it to return a double value rather than the default integer.

(int)Math.Ceiling((double)(System.DateTime.DaysInMonth(2009, 1) / 7));

As you can see I need the division to return a double so I can use the ceiling function.

Answer 1

A division of two int numbers returns an int, truncating any decimal points. This is generally true for other data types as well: arithmetic operations don't change the type of their operands.

To enforce a certain return type, you must therefore convert the operands appropriately. In your particular case, it's actually sufficient to convert one of the operators to double: that way, C# will perform the conversion for the other operand automatically.

You've got the choice: You can explicitly convert either operand. However, since the second operand is a literal, it's better just to make that literal the correct type directly.

This can either be done using a type suffix (d in the case of double) or to write a decimal point behind it. The latter way is generally preferred. in your case:

(int)Math.Ceiling(System.DateTime.DaysInMonth(2009, 1) / 7.0);

Notice that this decimal point notation always yields a double. To make a float, you need to use its type suffix: 7f.

This behaviour of the fundamental operators is the same for nearly all languages out there, by the way. One notable exception: VB, where the division operator generally yields a Double. There's a special integer division operator (\) if that conversion is not desired. Another exception concerns C++ in a weird way: the difference between two pointers of the same type is a ptrdiff_t. This makes sense but it breaks the schema that an operator always yields the same type as its operands. In particular, subtracting two unsigned int does not yield a signed int.