Re: [Dev-C++] numerical error in a simple program

[w.y. <yok...@ve...>]

"Luís Brás" wrote:
> 
>  a = 0.333300000000000
>  a = 3.333000000000000
>  a = 33.329999999999998
>  a = 333.299999999999950
>  a = 3332.999999999999500

> How to avoid this error?

You have to design your own floating point format type, or find a
library that does it...
If you need _exact_ FP calculations, the built in types can't do it.