a****o 发帖数: 42 | 1 man (well i suppose): MATLAB gives the perfct answer for you.
some observation justify this:
1. first xxe-16 is a number very small (very close to smallest number
in MATLAB), so the answer is accept for YOUR problem, or
better to say it is acceptable for the specific algorithm
you are using.
Actually you are showing the number in short format, if you show numbers
in long format, you will know that there is no perfect number.
2. Though your problem seems to be very easy, it is in fact not
a trivi | o********g 发帖数: 162 | 2 it is e-16, not e16, why did you call it wired? hard to believe?
matlab uses very accurate inside representation, while in normal format
use less digitals to display. In this case, the non-zero e is simply
because of round error. |
|