In MATLAB, due to finite precision, if you start with 0.1 and add 0.1 nine times, the result is slightly less than 1.
In Maple, 0.1 is represented as an sfloat, a software float, which uses a base 10 representation and so is able to represent 0.1 as exactly 1 * 10^(-1) . Adding 9 copies of that to the initial 0.1 gives you exactly 1 as a result.
Therefore the MATLAB version that madhan ravi suggests gives one more iteration than the Maple version does.
I have enclosed the translation of the Maple code to MATLAB code. There are, however, a few small differences in formatting, and I took the short-cut of implementing x^2 as x*x
It would have been a lot simpler and faster to make a minor modification to the algorithm to not add up 0.1's instead of having to implement decimal arithmetic in order to do exactly the same thing that Maple does.
