Numeric Data Types

• The integer (int) datatype
  • This represents a subset of the integers.
  • Probably -2³¹ to 2³¹-1
  • Or -2⁶³ to 2⁶³-1
  • Or -2¹⁵ to 2¹⁵-1
  • The C/C++ standard does not give a size, that is determined by the hardware/compiler implementer.
  • operations include +,-,*,/ and %
    • % is modulus.
    • It gives the remainder
    • 5 % 1 = 0 because 5 = 1 * 5 + 0
    • 5 % 2 = 1 because 5 = 2 * 2 + 1
    • 5 % 3 = 2 because 5 = 3 * 1 + 2
    • 5 % 4 = 1 because 5 = 4 * 1 + 1
    • This turns out to be very useful
  • If we try to compute a number that is larger than can be stored, integer overflow occurs.
  • Input, Output
    • cin >> intVar
    • Skips all white space
    • Reads as much of an integer as it can
    • Stops when it hits a noninteger
    • stores the results in the variable
  • Be careful, if an integer starts with a 0, it is assumed to be in octal.
  • This is just one of a family of integer types.
    • The char type is actually an integer.
      • This is the smallest integer class.
      • I/O on char is different.
      • It is not common to use a char as an integer type.
    • short is the next class
      • It is at least as big as a char, but may be larger.
    • int is next
      • It is at least as big as a short but may be larger.
    • long is next
      • It is at least as big as a int but may be larger.
      • This would be used for computing with large numbers.
  • Each of these may be modified by unsigned
    • This means only zero and positive values.
    • The range is shifted so that it starts at 0.
    • And goes through 2ⁿ -1
    • It is rare to use an unsigned
  • It is rare to use anything other than an int.
• Reading Exercise: change.C
• The floating point types
  • Floating point numbers are numbers in scientific notation.
    • 1.23x10¹⁰⁰ is a very large number.
    • 1.23x10^-100 is a very small number.
    • 1.23x10⁰ is a number very close to 1.
  • float is the most common type.
  • but double and long double are also available.
  • The later two will be much slower to compute with.
  • But will provide more accuracy.
  • Floating point constants can be
    • 1
    • 1.3
    • -1.342
    • 0.000002345
    • 3E5 -> 3x10⁵
    • 4.234E-12 -> 4.234x10^-12
  • Floating point numbers may not be accurately represented by the computer.
  • More on this later.
  • I/O of floats
    • Input: as with ints, try to read as much of the data from the stream as possible.
    • Output, more details at the end of the chapter.
• Expressions with numeric data types
  • For the most part, equations are as you expect.
  • But the right hand side of the equation is evaluated and assigned to the left hand side.
  • This is different from math where the two sides are the same.
  • The order of operations is as you learned, and is listed in appdendix B.
  • We use () as you would expect to change the order of operations.
• Reading Exercise: burger.C