Thanks - yes, classical interface, I should have mentioned that. But no, of course I did not want to read thousands of lines :-) That example was through a typo (forgot some brackets) and it needed some time to guess the reason (as I had not expected such a long result).

Hi, I looked at your statement and would like to see some stuff about multi-dim integration - is there Maple code available?

what I miss is: input a fct and get back a formal sum for the Taylor series (and the radius of convergence if possible) while for 'basic' functions that can be done through the FunctionAdvisor it not clear for me how - for example - to get 1/((1-s)^(1/2)) = Sum( (2*k)!/(k!)^2*2^(-2*k)*s^k, k=0..infinity)

what I miss is: input a fct and get back a formal sum for the Taylor series (and the radius of convergence if possible) while for 'basic' functions that can be done through the FunctionAdvisor it not clear for me how - for example - to get 1/((1-s)^(1/2)) = Sum( (2*k)!/(k!)^2*2^(-2*k)*s^k, k=0..infinity)

You can write down the symbolic taylor series by summing to infinity like f(x) = f(0) + Sum( (D@@i)(f)(x) / i!, i=1 .. infinity ) or in some point a different from 0. What I miss are specific presentations as series (in Sum form) for concrete functions as input (as I can get it from FunctionAdvisor).

You can write down the symbolic taylor series by summing to infinity like f(x) = f(0) + Sum( (D@@i)(f)(x) / i!, i=1 .. infinity ) or in some point a different from 0. What I miss are specific presentations as series (in Sum form) for concrete functions as input (as I can get it from FunctionAdvisor).

Besides the one or other thing there is what I really miss (or I am too stupid to handle that ...): A good search functionality and a reasonable way to save threads to my local disk. While the first should be possible, the second may be not that easy (currently I print them as postscript, which is ugly for a web based source [and it _is_ ugly to look at the results ...]). It would be great (no, not with graphics contained, just slim). And ... well, a forum with so much contributions is not really beta, it is alive :-)

Hm ... I do not speak Italian (?) you may get answers, if you post, what is the hardware and software for your PC (which machine, operating system etc). When I have such problems, I would first de-install Maple. Then install it again. What I meant is: use a directory name like Maple10, not Maple 10 (with a blank space between Maple and 10). For Java: I think Maple has its own version 1.4.2, do not touch that (but it should not matter if you have any Java installed, just do nothing within Maples directories). But I can not say more, as the Java interface does not run on my old PC, I use the classical interface. If you have an old PC (using WinME or similar) you occasionally will have trouble with modern software ... The best would be: ask, where you have bought your Maple, they should be able to explain it in your language.

Hm ... I do not speak Italian (?) you may get answers, if you post, what is the hardware and software for your PC (which machine, operating system etc). When I have such problems, I would first de-install Maple. Then install it again. What I meant is: use a directory name like Maple10, not Maple 10 (with a blank space between Maple and 10). For Java: I think Maple has its own version 1.4.2, do not touch that (but it should not matter if you have any Java installed, just do nothing within Maples directories). But I can not say more, as the Java interface does not run on my old PC, I use the classical interface. If you have an old PC (using WinME or similar) you occasionally will have trouble with modern software ... The best would be: ask, where you have bought your Maple, they should be able to explain it in your language.

Given a variable char* x or INTEGER8* x: how do i convert that to a Maple object Array(datatype = (integer)[1],(order) = (C_order)) to be used from OpenMaple (by calling a Maple procedure with that)? Sorry if the Q is stupid, but I can not find it by playing with some "generated wrappers" (or missed it in the Guide).

Given a variable char* x or INTEGER8* x: how do i convert that to a Maple object Array(datatype = (integer)[1],(order) = (C_order)) to be used from OpenMaple (by calling a Maple procedure with that)? Sorry if the Q is stupid, but I can not find it by playing with some "generated wrappers" (or missed it in the Guide).

I have similar problems (Maple10 on WinME with classical interface), kernel mode = parallel).

Then i go to an empty line, say the last, type in one blank, copy and paste it in the same place ... stupid, but that helps

May be that is a step to it, but not elegant and I have not checked details (assumed that w0 < r, anti-derivative continous?, numerical examples to see it might be correct, branch cuts for the roots): Int(1/(1+(r-r0)^2/A^2)*(1/sqrt(1-w0^2/r^2)),r=w0..infinity); integrand(%); simplify(%) assuming w0 < r; # really ? Now use use a simple version of that result with r < 0 or 0 < r: theIntegrand:=r/(r^2-w0^2)^(1/2)/(B+r^2-2*r*r0+r0^2); int(theIntegrand,r)'; %: simplify(%): applyrule((r0^2-2*r0*(-B)^(1/2)-B-w0^2)^(1/2)=C1,%): applyrule((r0^2+2*r0*(-B)^(1/2)-B-w0^2)^(1/2)=C2,%): antideriv:=%; Then evaluate in the limits (which seems to exist) to get a result by replacing C1, C2 (the 'double root' may be a complex root?).

Thanks - not fine, but at least good to know ... No problem here for me, as the integrand is a Gauss bell multiplied by a positive function of known bound, so one can find a cut off and roughly even could estimate the error for that (or switch to external routines).

Thanks - not fine, but at least good to know ... No problem here for me, as the integrand is a Gauss bell multiplied by a positive function of known bound, so one can find a cut off and roughly even could estimate the error for that (or switch to external routines).