CIS3355: Business Data Structures
Fall, 2008
 

What if I need greater precision or magnitude ??

The 32-bit layout for real numbers we have been discussing is sometimes referred to as:

Single precision real numbers

That is because there is something called :

Double precision real numbers

Take a wild guess as to how many bits are involved.

Ah, 64 ???

Brilliant!!

How are the bits distributed between the mantissa and characteristic of the exponent for these number ???

That depends on the software package. For example, ANSI (The American National Standards Institute) requires that the data type double (in C/C++) have at least 10-decimal point of precision.

What if I need more precision ???

There is something called :

Long Double precision real numbers

Take a wild guess as to how many bits are involved.

Ah, 128 ???

Brilliant!!

What level of precision can I expect ???

ANSI ( requires that the data type long double (in C/C++) have at least 10-decimal point of precision.

Does that mean I can't get greater precision or greater magnitude ???

Sure you can, but you have to buy the specific software package that will allow you to do that. Check the manufacturers manuals.

How can we summarize all of the basic data types?