a more reasonable search functionality (as I have seen in other forums) would allow to search for authors and more ...

symbolical:

  [op(aa)];
  map( _x -> int(_x,t=-Pi/4..Pi/4), %); `+`(op(%)); evalf(%);

                        [0, 0, 0, 0, 0, 0, 0]

                                  0

numerical:

  [op(aa)]:
  map( _x -> evalf(Int(_x,t=-Pi/4..Pi/4)), %); `+`(op(%));

  [1.7355011477181, -0.90246059681647, -0.97188064272214, 0.69420045908582, 0., 0., 0.]

                           0.55536036726531

or

  Int(a,t=-u..u); value(%);
  limit(%,u=Pi/4, left); evalf(%);


                                1/2
                               2    Pi
                               -------
                                 16


                           0.27768018363490

a good choice, I like their contributions!

Being neither from Physics nor from Engineering I looked at his second sheet: a nice idea for an initial approximation of the inverse error function trying erf(x) ~ tanh(a*x), optimize for a in the range of interest (through the L2 norm, then differentiate w.r.t. a and solve numerical), then simply use arctanh to invert.

yes ... sorry for my sloppiness :-(

yes ... sorry for my sloppiness :-(

cool, yes ... thx for pointing out!

cool, yes ... thx for pointing out!

what about

 `^`(a,b) 

?

what about

 `^`(a,b) 

?

hm ... do not know, never carried around waitresses ... (no ... do not know the record ...) have not been at the fest for dozend years, not so much my thing the nicer and more typical places in Bavaria are smaller gardens where one aims to relax (and you almost never will meet drunken ones there, they mostly will not any more alcohol ...), one even can bring ones own bread, bacon and cheese with you

this is a great answer! whenever come to Munich ... let's have a beer (or better) ...

i let the older ones installed (one has to care for the Watcom compiler a bit if i remember correctly) and opening a file through double klick takes the latest version, but through the context menu (as Windows standard) one can even choose the versions

but in the cited example Maple can be of help to guess the correct statement: x/(x + floor(x)) + x/(x+frac(x)) -5/2; plot(%,x=0 ..6); or (the transformed one) x/(x+b) + x/(x+a)-3/2; subs(a=frac(x),b=floor(x),%); plot(%,x=-3 ..16);

may be it is also a personal taste, but i find that 'standard' disgusting (sorry for the harsh word) - it even prevents me to use most of the downloadable files when i considered switched from occasionally using Mupad to Maple and that standard would have been the only one - i would never had done it so i like using Maple - i really HATE that nerve-shattering interface (nicer looking, yes ... but how about to fresh up the real interface?) ok, enough ranting ... time for a beer :-)