dolfin team mailing list archive
-
dolfin team
-
Mailing list archive
-
Message #12252
Re: A minimal c++ Function test and some bugs
On Tue, Feb 17, 2009 at 05:05:16PM +0000, A Navaei wrote:
> 2009/2/17 Anders Logg <logg@xxxxxxxxx>:
> >> MyFunction f_(f);
> >
> > Here you copy the function f to f_ and since f has a zero vector, f_
> > will also have a zero vector.
>
> It seems when we create f in python, it doesn't come with a zero
> vector. That's why interpolate.py works.
>
> >
> >> f.interpolate(f_.vector(), f.function_space());
> >
> > Here f_ already has a zero vector or would get it by the call to
> > vector() if it didn't so again the result will be zero.
>
> So the work-around doesn't work in c++ (but for whatever reason it
> works in python). I would call the zero vector problem a bug.
I'd call it a feature.
--
Anders
Attachment:
signature.asc
Description: Digital signature
References
-
Re: A minimal c++ Function test and some bugs
From: Anders Logg, 2009-02-17
-
Re: A minimal c++ Function test and some bugs
From: A Navaei, 2009-02-17
-
Re: A minimal c++ Function test and some bugs
From: Anders Logg, 2009-02-17
-
Re: A minimal c++ Function test and some bugs
From: A Navaei, 2009-02-17
-
Re: A minimal c++ Function test and some bugs
From: A Navaei, 2009-02-17
-
Re: A minimal c++ Function test and some bugs
From: A Navaei, 2009-02-17
-
Re: A minimal c++ Function test and some bugs
From: Anders Logg, 2009-02-17
-
Re: A minimal c++ Function test and some bugs
From: A Navaei, 2009-02-17
-
Re: A minimal c++ Function test and some bugs
From: Anders Logg, 2009-02-17
-
Re: A minimal c++ Function test and some bugs
From: A Navaei, 2009-02-17