MathGroup Archive 2005

[Date Index] [Thread Index] [Author Index]

Search the Archive

Re: Converting between Spherical and Cartesian coordinates

  • To: mathgroup at smc.vnet.net
  • Subject: [mg53336] Re: Converting between Spherical and Cartesian coordinates
  • From: Alain Cochard <alain at geophysik.uni-muenchen.de>
  • Date: Thu, 6 Jan 2005 22:00:37 -0500 (EST)
  • References: <crg14d$bg8$1@smc.vnet.net> <crir73$s5k$1@smc.vnet.net>
  • Sender: owner-wri-mathgroup at wolfram.com

Alain Cochard writes:

 >  > I use spherical coordinates:
 >  > 
 >  >      In[1]:= << Calculus`VectorAnalysis` ; 
 >  > 
 >  >      In[2]:= SetCoordinates[Spherical]
 >  > 
 >  >      Out[2]= Spherical[Rr, Ttheta, Pphi]
 >  > 
 >  > and I convert:
 >  > 
 >  >      In[3]:= CoordinatesFromCartesian[{x,y,z}]
 >  > 
 >  > 		    2    2    2                  z
 >  >      Out[3]= {Sqrt[x  + y  + z ], ArcCos[------------------], ArcTan[x, y]}
 >  > 					       2    2    2
 >  > 					 Sqrt[x  + y  + z ]
 >  > 
 >  > Now, if I use the expressions for x, y, and z, I expect to get back to
 >  > Rr, Ttheta, Pphi, but:
 >  > 
 >  >      In[4]:= FullSimplify[%3/.{
 >  > 			  x->Rr Sin[Ttheta]Cos[Pphi],
 >  > 			  y->Rr Sin[Ttheta]Sin[Pphi],
 >  > 			  z->Rr Cos[Ttheta]}, 
 >  > 		  Rr>=0 && 0 <= Ttheta <= Pi && -Pi < Pphi <= Pi]
 >  > 
 >  >      Out[4]= {Rr, Ttheta, ArcTan[Rr Cos[Pphi] Sin[Ttheta], 
 >  > 
 >  >      >     Rr Sin[Pphi] Sin[Ttheta]]}
 >  > 
 >  > 
 >  > Even if I remove the equalities from the assumptions, I am not quite
 >  > there: 
 >  > 
 >  >      In[5]:= FullSimplify[%3/.{
 >  > 			  x->Rr Sin[Ttheta]Cos[Pphi],
 >  > 			  y->Rr Sin[Ttheta]Sin[Pphi],
 >  > 			  z->Rr Cos[Ttheta]}, 
 >  > 		  Rr>0 && 0 < Ttheta < Pi && -Pi < Pphi < Pi]
 >  > 
 >  >      Out[5]= {Rr, Ttheta, ArcTan[Cos[Pphi], Sin[Pphi]]}
 >  > 
 >  > But I would have thought that for a given Pphi in (-Pi,Pi), there is a
 >  > unique value for ArcTan[Cos[Pphi], Sin[Pphi]], which is... Pphi.  And
 >  > this seems to me consistent with the fact that a given (single) point
 >  > is unique, or, in other words, that to a given (perhaps
 >  > not-too-specific) {x,y,z} corresponds a unique {Rr, Ttheta, Pphi}, and
 >  > vice versa.
 >  > 
 >  > What am I missing here?
 >  > 
 >  > Thanks in advance,
 >  > AC
 >  > 
 > 

Astanoff writes:

 > Alain,
 > 
 > Seems that it does simplify if Pphi (as a latitude) ranges
 > from -Pi/2 to Pi/2 and if you don't allow equalities in coordinate
 > ranges :

Yes, of course, and ...


 > In[1]:=<<Calculus`VectorAnalysis`
 > 
 > In[2]:=SetCoordinates[Spherical]
 > Out[2]=Spherical[Rr,Ttheta,Pphi]
 > 
 > In[3]:=
 > Unprotect[CoordinateRanges];
 > CoordinateRanges[ ]={0<Rr<Infinity, 0<Ttheta<Pi, -Pi/2<Pphi<Pi/2};
 > Protect[CoordinateRanges];
 > 
 > In[6]:=sph=CoordinatesFromCartesian[{x,y,z}]
 > Out[6]={Sqrt[x^2 + y^2 + z^2], ArcCos[z/Sqrt[x^2 + y^2 + z^2]],
 > ArcTan[x, y]}
 > 
 > In[7]:=
 > car=sph /. Thread[{x,y,z} -> CoordinatesToCartesian[{Rr,Ttheta,Pphi}]]
 > Out[7]=
 > {Sqrt[Rr^2*Cos[Ttheta]^2 + Rr^2*Cos[Pphi]^2*Sin[Ttheta]^2 +
 > Rr^2*Sin[Pphi]^2*Sin[Ttheta]^2],
 > ArcCos[(Rr*Cos[Ttheta])/Sqrt[Rr^2*Cos[Ttheta]^2 +
 > Rr^2*Cos[Pphi]^2*Sin[Ttheta]^2 +
 > Rr^2*Sin[Pphi]^2*Sin[Ttheta]^2]],
 > ArcTan[Rr*Cos[Pphi]*Sin[Ttheta], Rr*Sin[Pphi]*Sin[Ttheta]]}
 > 
 > In[8]:=car // Simplify[#,CoordinateRanges[ ]]&
 > Out[8]={Rr,Ttheta,Pphi}

... and there is not even the need to redefine the range as you do
above:

    In[1]:= << Calculus`VectorAnalysis`;
    << Calculus`VectorAnalysis`;

    In[2]:= SetCoordinates[Spherical]

    Out[2]= Spherical[Rr, Ttheta, Pphi]

    In[3]:= CoordinatesFromCartesian[{x,y,z}]

		   2    2    2                  z
    Out[3]= {Sqrt[x  + y  + z ], ArcCos[------------------], ArcTan[x, y]}
					      2    2    2
					Sqrt[x  + y  + z ]

    In[4]:= FullSimplify[%3/.{
			      x->Rr Sin[Ttheta]Cos[Pphi],
			      y->Rr Sin[Ttheta]Sin[Pphi],
			      z->Rr Cos[Ttheta]}, 
		      Rr>0 && 0 < Ttheta < Pi && -Pi/2 < Pphi < Pi/2]

    Out[4]= {Rr, Ttheta, Pphi}

but with this restricted range one does not cover the whole space.  So
my initial question remains...

Thanks anyway,
Alain


  • Prev by Date: logical indexing using more than one variable in a pure function
  • Next by Date: Re: FullSimplify with Assumptions
  • Previous by thread: Re: Converting between Spherical and Cartesian coordinates
  • Next by thread: Re: Converting between Spherical and Cartesian coordinates