Re: Integrate fails without warning?

*To*: mathgroup at smc.vnet.net*Subject*: [mg35283] Re: Integrate fails without warning?*From*: "Raf" <r_a_f at yahoo.it>*Date*: Sat, 6 Jul 2002 05:44:30 -0400 (EDT)*References*: <ag3ef9$72v$1@smc.vnet.net>*Sender*: owner-wri-mathgroup at wolfram.com

Juan, I know it's possible to obtain the right answer in many way, but what surprise me is that: 1) the result is WRONG; and... 2) no error messages are produced; and... 3) no warning messages are produced. In mathematica-help i didn't find any notes about the functions that must not be used in Integrate; so i call this a BUG. Regards, Bye, Raf. "Juan Fuentes" <jfuentes at ff.oc.uh.cu> ha scritto nel messaggio news:ag3ef9$72v$1 at smc.vnet.net... > Raf > I did the integration in another way and obtaine the correct value > s[t_] := UnitStep[0.5 - t] - UnitStep[t - 0.5] > Integrate[s[t], {t, 0, 1}] > 0. > Maybe the problem is in the function. > With best regards > Juan > > On Wed, 3 Jul 2002, Raf wrote: > > > This regards not only Calculus`Integration`, but also the built-in > > Integrate: > > Defining a square wave... > > In[1]:= ondQua[t_] := If[OddQ[Floor[2t]], -1, 1] > > ....Mathematica fails completely in computing its defined integral... > > In[2]:= Integrate[ondQua[t], {t, 0, 1}] > > Out[2]= 1 > > ....without any warning or error message !!! > > How is it possible? > > > > With NIntegrate the things goes better, but i believe that a similar risk > > with Integrate would have to be documented somewhere; if exists some notes > > about this using of Integrate on Mathematica-help, please tell me where. > > > > Waiting your answers, > > Thanks, > > Raf. > > > > > > > > > >