Re: Same .m file: different results with different versions of Octave

[Thomas D. Dean]

On Mon, 2010-04-19 at 12:00 -0700, Søren Hauberg wrote:

> which just doesn't make sense to me. In the specific case where the
> infinite numbers came from 'sinh' and 'cosh' it might make mathematical
> sense, but in the general setting I don't see how the above equation
> could ever hold.

My claim is:

    limit(tanh(n*(1+i)),n=infinity) == 1

Note that
    limit(sinh(n*(1+i)),n=infinity) == exp(i)*infinity
    limit(cosh(n*(1+i)),n=infinity) == exp(i)*infinity
    

    limit(sinh(n*(1+i)) - cosh(n*(1+i)), n = x) is -exp(-x)/(exp(i)^x)
which rapidly goes toward 0+0i.

But, this is not the point.  tanh is the problem I was discussing.