Re: eval() for user-defined matrices?

[Carlo de Falco]

On 9/7/06, the_verge <address@hidden> wrote:
> Hi all,
>
> Is there a way to implement eval("thing_x") if thing_x is a vector?  Or is
> there some alternative command that will do the same thing?  For example:
>
> thing_x = [1 3 6 9 0];
> eval("thing_x")
>
> returns "invalid vector index = 116"
> instead of:
>
> thing_x = 1 3 6 9 0
>
> Any ideas on how to do this?
> Note:  I would like class(eval(thing_x)) to be double
>
> Thanks,
> Vergil
> --
> View this message in context: 
> http://www.nabble.com/eval%28%29-for-user-defined-matrices--tf2230998.html#a6183851
> Sent from the Octave - General forum at Nabble.com.


If you define

> > thing_x = [1 3 6 9 0];

> > thing_x

thing_x =

 1  3  6  9  0

> > class (thing_x)
ans = double

you don't need to eval(), thing_x is already a vector of double...
what eval() does is

1) take _a_string_ as input
2) interpret the octave commands contained in the string
3) return the final result of the calculation as output

so you probably wanted to do:

> > thing_x = "[1 3 6 9 0]";
> > class(thing_x)
ans = char

> > eval(thing_x)
ans =

 1  3  6  9  0
> > class(ans)
ans = double

-
c.