precision in C

[EvLer]

Hello everyone,

I need to be able to limit(or control) precision of double datatype to 10 decimal places and not more or less. How would i do it? I can also use float or any other datatype, the issue is making precision only 10 decimal places:

0.0123456789

and this is not output formatting.... I am working with a program that does a scientific computation, so precision is important in calculations, not for the display.

Thank you in advance.

 

[xwb]

Guess you need to work in a language like Coral66, Ada or Cobol where the number of decimal places can be fixed.

Assuming fmod behaves the same way as % and that they haven't done it properly for -ve numbers, after each calculation, you could do something like

x -= fmod (x, 1E-10);

Very expensive though.

 

[EvLer]

thank you for the idea! it works!

 

[WalterContrata]

Neat!

EvLer,

If the number is greater than or equal to 1 ( or less than 0.1) how should it be rounded? For example, does

1.12345123451 get rounded to
1.1234512345 (10 digits after decimal) or
1.123451234 (10 significant figures) ?