Fix some warnings by adding explicit typecasts

[tsparber]

If more warnings are enable in Xcode these warnings are shown:

• Implicit conversion turns floating-point number into integer: 'float' to 'size_t' (aka 'unsigned long’)
• Implicit conversion loses floating-point precision: 'double' to 'CGFloat' (aka 'float’)

This commit fixes these by adding explicit typecasts.