Re: [changeset] Polyfit with scaling

[Ben Abbott]

On Feb 20, 2008, at 1:17 AM, John W. Eaton wrote:

> On 20-Feb-2008, John W. Eaton wrote:
>
> | I applied this changeset with a few minor style changes.
>
> Now I see the following error when running make check:
>
> > > > > > processing /export/home/jwe/src/octave/scripts/polynomial/ 
> > > > > > polyfit.m
>    ***** test
>   x = [ -1196.4, -1195.2, -1194, -1192.8, -1191.6, -1190.4, -1189.2,  
> -1188, \
>          -1186.8, -1185.6, -1184.4, -1183.2, -1182];
>   y = [ 315571.7086, 315575.9618, 315579.4195, 315582.6206,  
> 315585.4966,    \
> 	 315588.3172, 315590.9326, 315593.5934, 315596.0455,  
> 315598.4201,    \
>          315600.7143, 315602.9508, 315605.1765 ];
>   [p1, s1] = polyfit (x, y, 10);
>   [p2, s2, mu] = polyfit (x, y, 10);
>   assert (s1.normr, 0.11264, 0.1)
>   assert (s2.normr < s1.normr)
>  !!!!! test failed
>  error: assert (s1.normr,0.11264,0.1) expected
>   0.11264
>  but got
>   0.87635
>  maximum absolute error 0.763709 exceeds tolerance 0.1
>
> Any clues?
>
> jwe

My bad.  I should have tested for a relative improvement over the  
unnormalized case. I suspect differences in lapack/blas produced the  
error (it passes on my Mac)

In any event, I've modified the tests to respect a relative comparison.

Ben

polyfit.changeset
Description: Binary data