numer and float -- was RE: exponential with complex exponents

I think the idea was that ev(...,numer)  would run some math functions like
exponential to be evaluated, if their arguments were numbers.
 The result is then simplified. Ev(...,float)  would change rational
constants like 1/2 to floating point. 

So far as I can tell,  using float(..) on an object that is not a rational
number or integer or a constant is not specified.  In particular, applying
it to a complicated expression is a mistake.

Use ev(..., float) if you want ALL the subexpressions to be floated.

 And that should probably be replaced to distinguish single, double, bfloat
etc.  e.g.  
Ev(..., dfloat)  .



> -----Original Message-----
> From: maxima-bounces at math.utexas.edu [mailto:maxima-
> bounces at math.utexas.edu] On Behalf Of Barton Willis
> Sent: Wednesday, September 20, 2006 3:33 PM
> To: Joel B. Mohler
> Cc: maxima at math.utexas.edu
> Subject: Re: [Maxima] exponential with complex exponents
> 
> And yet another way (sigh...)
> 
> (%i11) e : %e^(10/21*%pi*%i);
> (%o11) %e^((10*%i*%pi)/21)
> (%i12) e,numer;
> (%o12) 0.99720379718118013*%i+0.074730093586424393
> (%i13) float(e);
> (%o13) 2.7182818284590451^((10*%i*%pi)/21)
> 
> This is confusing. I don't know the history of numer and float,
> but we should just *one* way (that always works!) to convert
> constant expressions to a float. We're always looking for new
> volunteers.
> 
> Barton
> _______________________________________________
> Maxima mailing list
> Maxima at math.utexas.edu
> http://www.math.utexas.edu/mailman/listinfo/maxima