Data Types

• Here is what we have so far
• 
• Integral Types
  • These are things that have no fractional parts.
  • Unless we tell it not to, all cases support minus sign.
  • prefacing a type with unsigned tells it not to use a minus sign.
    • This gives us an additional bit in the representation.
    • There is some protection here, but look consider:

      main () {
         unsigned int x;
      
         x = -1;
      
         cout << "X is " << x << endl;
      }
      		   

      which produces

      [bennett@douc54 three]$ g++ ints.cpp
      ints.cpp: In function `int main ()':
      ints.cpp:6: warning: assignment of negative value `-1' to `unsigned 
      int'
      ints.cpp:6: warning: argument of negative value `-1' to `unsigned int'
      [bennett@douc54 three]$ a.out
      X is 4294967295
      
      

    • A warning is a message from the compiler that you are doing something wrong, but it can still compile the code.
    • You should eliminate warnings from your code.
    • You should never submit a program to be graded that compiles with warnings.
  • Another note, literals beginning with 0 are assumed to be in octal or base 8.

    main () {
      int foo;
    
      foo = 010;
      cout << "Foo is "<< foo << endl;
    }
    
    Will produce
    Foo is 8
    	  

  • Sizes of these types depends on the implementation (or hardware).
  • The sizeof function will return the size of a data type.
  • Look at this program that prints sizes of the standard types.
  • This program produces (on my machine).

    
    An bool is        1 bytes 
    An char is        1 bytes 
    An short is       2 bytes 
    An int is         4 bytes 
    An long is        4 bytes 
    An float is       4 bytes 
    An double is      8 bytes 
    An long double is 12 bytes 
    	 

  • It is possible to get a number that is too large to store, this is called overflow.

    
    main () {
       int pow;
    
       pow = 2;
       cout << "Pow is " << pow << endl;
       pow = pow * pow;  // two squared
       cout << "Pow is " << pow << endl;
       pow = pow * pow;  // two to the eighth
       cout << "Pow is " << pow << endl;
       pow = pow * pow;   // two to the sixteenth
       cout << "Pow is " << pow << endl;
       pow = pow * pow;   // two to the thirtysecond
       cout << "Pow is " << pow << endl;
       pow = pow * pow;   // two to the 64
       cout << "Pow is " << pow << endl;
    
    }
    
    Will produce 
    Pow is 2
    Pow is 4
    Pow is 16
    Pow is 256
    Pow is 65536
    Pow is 0
    
    	 

• Floating Point types
  • Hold numbers with a decimal point.
  • Sizes vary with compiler and hardware.
  • Scientific notation is supported.

    int main () {
       float x;
    
       x = 1e3;
       cout << x << endl;
    }
    
    Will produce
    
    1000
          

  • 1e3 is 1x10^3
  • Real numbers are not represented exactaly.
    • Some round off errors may occur.
    • 1/3 = .333333333333..., infinate number of 3s
• How do we declare {float,int} constants?
• How do we declare {float, int} variables?
• Simple Arithmetic
  • +, -
    • binary operations
      • day = day + 1;
      • newcount = oldcount - currentcount;
    • unary operations
      • sum = -1;
      • total = +price;
  • *
    • as +,- but only as a binary operator
  • /
    • binary operator.
    • As normal on floats
    • 10/4 = 2 in integers.
    • 10.0/4.0 = 2.5 in floats.
    • x/0 is not permitted, no divison by zero
  • % - modulo or integer remainder.
    • Remainder operation in division
    • 10%4 = 2
    • Only for integers.
    • x%0 is not permitted, no divison by zero
• Please consider this example program.
  • This program is also on cslab41.
  • To copy it to your directory type
  • cp ~dbennett/130/ch_3/payday.cpp .