Re: Real/Comlex function problem with D
- To: mathgroup at smc.vnet.net
- Subject: [mg74240] Re: Real/Comlex function problem with D
- From: dh <dh at metrohm.ch>
- Date: Thu, 15 Mar 2007 05:02:18 -0500 (EST)
- References: <et8e29$pqb$1@smc.vnet.net>
Hi, Consider ff1[x_]:=D[ff[x],x]. Now f[3.] evaluates to D[ff[3]],3], hardly what you want. Presumably your mean ff1[x]:=ff'[x]. But now we have the problem that we need the derivative of Abs and in general (in the complex field) this derivative is not defined. You could try to make a definition for Element[x,Reals], but you need to take care of branch cuts, that depend on your F. What you can do is to approximate the real Arg[x] with real x by e.g. Interpolation or any other approximation metod like e.g.:, Calculus`Pade` , NumericalMath`Approximations` , NumericalMath`NSeries` , NumericalMath`PolynomialFit` Daniel bar at ANTYSPAM.ap.krakow.pl wrote: > Hi, > I have a complex function F[x] > When i defined: > ff[x_]:=Arg[F[x]] (* i tried ComplexExpand[Arg[F[x]]] too *) > > I obtainned good Plot[ff[x]] - its real - (no complex) > > When i try > ff1[x_]:=D[ff[x],x]; > I found ff1 is Complex funtion again !! > > Why ? > > (i tried Evaluate; Element[x,Reals]; FullSimplify - always the same) > > Regards, Olaf > >