Commodore 128 Programmer's Reference Manual page 23

n
n
n
n
n
n
n
BASIC BUILDING BLOCKS AND BASIC 7.0 ENCYCLOPEDIA 13
which the contents vary or change throughout the course of a program. The last way to store
information is to use an array, a series of related memory locations consisting of variables.
Each of these three units of memory storage can have three different types of
information or data assigned. The three data types are INTEGER, FLOATING-POINT
or STRING. Integer data is numeric, whole number data—that is, numbers without
decimal points. Floating-point is numeric data including fractional parts indicated with a
decimal point. String data is a sequential series of alphanumeric letters, numbers and
symbols referred to as character strings. The following paragraphs describe these three
data types and the way each memory storage unit is assigned different data type values.
CONSTANTS: INTEGER,
FLOATING-POINT AND STRING
INTEGER CONSTANTS
The value assigned to a constant remains unchanged or constant throughout a program.
Integer constants can contain a positive or negative value ranging from -32768 through
+ 32767. If the plus sign is omitted, the C128 assumes that the integer is positive.
Integer constants do not contain commas or decimal points between digits. Leading
zeros are ignored. Integers are stored in memory as two-byte binary numbers, which
means a constant requires 16 bits or two bytes of memory to store the integer as a base
two number. The following are examples of integer constants:
1
1000
-32
0
-32767
FLOATING-POINT CONSTANTS
Floating-point constants contain fractional parts that are indicated by a decimal
point. They do not contain commas to separate digits. Floating-point constants may be
positive or negative. If the plus sign is omitted, it is assumed that the number is
positive. Again, leading zeros are unnecessary and ignored. Floating-point constants are
represented in two ways depending on their value:
1. Simple Number Notation
2. Scientific Notation
In simple number notation, the floating-point number is calculated to ten digits of
precision and stored using five bytes, but only nine digits are displayed on the screen or
printer. If the floating-point number is greater than nine digits, it is rounded according to
the tenth digit. If the tenth digit is greater than five, the ninth digit is rounded to the next
higher digit. If the tenth digit is less than five, the ninth digit is rounded to the next
lower digit. The rounding of floating-point numbers may become a factor when calculat-