Calculation Error sum float

Question

0 个投票

I Write This Code:

clc;
clear;
format long
a=1;
for n=1:1:100
    a=a+.1;
end
a

and answer is:

a =
    10.999999999999979

!!!!! Why not: a= 11

How I Can Solve it?

1 个评论
显示 -1更早的评论隐藏 -1更早的评论

James Tursa 2014-5-16

See Azzi's link below. But for your particular case this is the decimal equivalent of the closest number to 0.1 in IEEE double format:

>> num2strexact(0.1)
ans =
0.1000000000000000055511151231257827021181583404541015625

You can find num2strexact here:

Answer 1

Azzi Abdelmalek 2014-5-16

1 个投票

Maha dasra 2014-5-16

How I Can Solve it?

José-Luis 2014-5-16

Use a computer that doesn't use the binary system. There are a few such strange base 10 machines out there.
Try variable precision arithmetic. Either with the symbolic toolbox or the excellent vpa from the file exchange.