calculation error or double precision limit?

Hello,

I was wondering about the result when calculating following simple formula:

K>> 0.067 * 1e9
ans =
   6.7000e+07

Following gives the correct result (integer):

K>> 0.067 * 1e8 * 1e1
ans =
    67000000

Where is the difference in the calcualation?

Is it an error or a double precision limit?

Answer 1

[Stephen23]

You should download this and give it a try with the same examples as above:

http://www.mathworks.com/matlabcentral/fileexchange/22239-num2strexact--exact-version-of-num2str-

Also have a look at the intermediate values, such as 0.067 * 1e8. This shows that there are a few digits that creep in after the end of the double precision limit (fifteen digits).