Re: [ublas] [bindings][lapack] fortran logical_t [Re: cmputational/stegr problems]

Subject: Re: [ublas] [bindings][lapack] fortran logical_t [Re: cmputational/stegr problems]
From: Hidekazu Ikeno (hide.ikeno+boost_at_[hidden])
Date: 2009-04-15 04:51:45

Hi,

Fortran has four different LOGICAL types, LOGICAL(1), LOGICAL(2),
LOGICAL(4)
and  LOGICAL(8) which might be convertible to integral types with same byte
size
in C++ (for example, char [1 byte], short [2 byte], int [4 byte], long int
[8 byte]
for lp64 architecture).

By default, "LOGICAL" is same as "LOGICAL(4)", but it is possibly
"LOGICAL(8)" if special
option is specified at the compile time. LOGICAL is inconvertible to bool in
either case,
I think.

Best regards,

Hidekazu Ikeno

2009/4/15 Rutger ter Borg <rutger_at_[hidden]>

> Hidekazu Ikeno wrote:
> >
> > Related to stemr, a new logical argument tryac is required
> > (sorry, I didn't notice that argument before).
> > It doesn't work at the moment since logical_t is defined as void.
>
> Should logical_t be a bool in this case? Would be interesting to compare
> this with the logical_t used by gees, which (I think) should be something
> like a function pointer of a function accepting a complex number and
> returning a bool.
>
> Kind regards,
>
> Rutger ter Borg
>
>
>
>
> _______________________________________________
> ublas mailing list
> ublas_at_[hidden]
> http://lists.boost.org/mailman/listinfo.cgi/ublas
>