On 1/30/11 4:55 PM, Barton Willis wrote:
> This method requires insight:
>
> (%i20) factor(integrate((3*t+2)^5,t,-2/3,x));
> (%o20) (3*x+2)^6/18
>
> Sometime ago there were, I think, other suggestions for doing this on the list.
> I looked, but I wasn't able to find them...
>
Yes, this has come up before. For this particular integral (and others
like it), some routine has a magic 6 hardwired in it. If the exponent
is less than 6, the integrand is expanded before doing the integration.
If the exponent is >= 6, the integrand is not expanded and you would get
a result in terms of the integrand with one higher exponent.
Nothing came of the discussions because in both cases maxima was
correct, even if the answer may not have been what the user wanted. I
don't think anyone wanted to create yet another global magic variable to
control this behavior, and there was no consensus on what is the desired
answer.
Ray