Re: 3.0.3 release candidate

[Jaroslav Hajek]

On Wed, Sep 24, 2008 at 3:48 PM, John W. Eaton <address@hidden> wrote:
> On 24-Sep-2008, Michael Goffioul wrote:
>
> | On Wed, Sep 24, 2008 at 3:28 PM, John W. Eaton <address@hidden> wrote:
> | > | Compiles and runs fine under Windows/MSVC. Only 2 test failures
> | > | in interpft.m due to tolerance slightly exceeded.
> | >
> | > How much?  Perhaps we should increase the tolerance.
> |
> | Not much. In both cases:
> |
> | maximum absolute error 2.498e-016 exceeds tolerance 2.22045e-016
>
> OK, so how about the following change?
>

Applied to 3.0.x


> Thanks,
>
> jwe
>
>
> # HG changeset patch
> # User John W. Eaton <address@hidden>
> # Date 1222264078 14400
> # Node ID 877faa7aa9b2df9193425119df746d5a024c5783
> # Parent  177117c7e375d63229f9dbad221114e313aa99ea
> interpft.m: increase tolerance
>
> diff --git a/scripts/ChangeLog b/scripts/ChangeLog
> --- a/scripts/ChangeLog
> +++ b/scripts/ChangeLog
> @@ -1,3 +1,7 @@
> +2008-09-24  John W. Eaton  <address@hidden>
> +
> +       * general/interpft.m: Increase tolerance in tests.
> +
>  2008-09-23  Francesco Potorti`  <address@hidden>
>
>        * plot/pcolor.m: Improve doc string.
> diff --git a/scripts/general/interpft.m b/scripts/general/interpft.m
> --- a/scripts/general/interpft.m
> +++ b/scripts/general/interpft.m
> @@ -106,9 +106,9 @@
>
>  %!shared n,y
>  %! x = [0:10]'; y = sin(x); n = length (x);
> -%!assert (interpft(y, n), y, eps);
> -%!assert (interpft(y', n), y', eps);
> -%!assert (interpft([y,y],n), [y,y], eps);
> +%!assert (interpft(y, n), y, 2*eps);
> +%!assert (interpft(y', n), y', 2*eps);
> +%!assert (interpft([y,y],n), [y,y], 2*eps);
>
>  %!error (interpft(y,n,0))
>  %!error (interpft(y,[n,n]))
>
>



-- 
RNDr. Jaroslav Hajek
computing expert
Aeronautical Research and Test Institute (VZLU)
Prague, Czech Republic
url: www.highegg.matfyz.cz