Re: Debracketing array symbols
- To: mathgroup at smc.vnet.net
- Subject: [mg92119] Re: Debracketing array symbols
- From: magma <maderri2 at gmail.com>
- Date: Sat, 20 Sep 2008 05:01:16 -0400 (EDT)
- References: <gavqna$eon$1@smc.vnet.net> <gavt05$g4l$1@smc.vnet.net>
On Sep 19, 11:56 am, dh <d... at metrohm.ch> wrote:
> Hi carlos,
>
> you could e.g. try:
>
> {A[1],A[2],A[3],A[4]}/.x_[y_]:>Symbol[ToString[x]<>ToString[y]]
>
> hope this helps, Daniel
>
>
>
> car... at colorado.edu wrote:
> > Hi,
>
> > For a class project I generate 1D and 2D arrays of arbitrary size with
> > entries <letters>[i] and <letters>[i,j] that I need then to convert
> > to <letter>i and <letter>ij for export to other languages that do not
> > allow brackets in symbol names. E.g.
>
> > 1D: {A[1],A[2],A[3],A[4]} -> {A1,A2,A3,A4}
>
> > 2D: {{W[1,1],W[1,2]},{W[2,1],W[2,2]}} -> {{W11,W12},{W21,W22}}
>
> > Question: would this be the simplest way to do it?
>
> > DebracketArrayEntries[Wm_]:=Module[{dim,rep,i,j,n,m,s,Wr},
> > rep={"["->"","]"->"",","->""," "->""}; Wr=Wm;
> > dim=Length[Dimensions[Wm]];
> > If [dim==1,n=Length[Wm]; Print["n=",n];
> > For [i=1,i<=n,i++, s=ToString[Wm[[i]]];
> > Wr=Wr/.Wm[[i]]->Symbol[StringReplace[s,rep=
]] ]];
> > If [dim==2,{n,m}=Dimensions[Wm];
> > For [i=1,i<=n,i++, For[j=1,j<=m,j++, s=ToStri=
ng[Wm[[i,j]]];
> > Wr=Wr/.Wm[[i,j]]->Symbol[StringReplace[s,r=
ep]] ]]];
> > Return[Wr]];
>
> > The procedural style is to simplify conversion to C++.
>
> --
>
> Daniel Huber
>
> Metrohm Ltd.
>
> Oberdorfstr. 68
>
> CH-9100 Herisau
>
> Tel. +41 71 353 8585, Fax +41 71 353 8907
>
> E-Mail:<mailto:d... at metrohm.com>
>
> Internet:<http://www.metrohm.com>
This is a perfect showcase of the expressive power of Mathematica versus
traditional procedural programming.
Daniel Huber's one line solution using ReplaceAll simply replaces all
occurrences of something in form of x[y] with a customized new symbol.
Anybody can compare this with the proposed procedural code with all
its ifs and buts, but do not waste your whole weekend :-)
PS: Huber's code deals with the 1D case.
Going to 2D should be almost obvious for everybody at this point.