[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Problem with cor (was: Re: mkoctfile problems with Octave 3.4.0, OSX 10.
From: |
Julien Salort |
Subject: |
Problem with cor (was: Re: mkoctfile problems with Octave 3.4.0, OSX 10.6.7) |
Date: |
Thu, 21 Apr 2011 16:51:59 +0200 |
Le 21 avr. 2011 à 16:39, Vic Norton a écrit :
> There is a problem with the cor function in this version of Octave. I am
> attaching illustrative computations using this Octave-3.4.0 and and your
> previous Octave-3.5.0+.
How does it compare with previous 3.4.0 builds ? (built from the same sources)
ie. does it have something to do with the way I built the release or only with
some possible changes in cor/cov related M files ?
(the 3.5.0+ release is built using the latest sources from the mercurial tree.
I have no idea how many things may behave differently.)
If this is related to how I built the package, I have no idea what the problem
could be...
Any idea ?
> ###
> octave-3.4.0> cor(Ra)
> error: cov: X and Y must have the same number of observations
> error: called from:
> error:
> /Applications/Octave.app/Contents/Resources/share/octave/3.4.0/m/statistics/base/cov.m
> at line 104, column 7
> error:
> /Applications/Octave.app/Contents/Resources/share/octave/3.4.0/m/statistics/base/corrcoef.m
> at line 71, column 7
> error:
> /Applications/Octave.app/Contents/Resources/share/octave/3.4.0/m/statistics/base/cor.m
> at line 34, column 10
> octave-3.4.0> cov(Ra)
> ans =
>
> 1.0196 -1.2736 -1.9540 -1.0084 -1.0189
> -1.2736 7.1934 9.9016 7.3092 8.1031
> -1.9540 9.9016 15.1665 9.5205 10.9879
> -1.0084 7.3092 9.5205 8.9518 9.2089
> -1.0189 8.1031 10.9879 9.2089 11.5935
> octave-3.4.0> cov(Ra, 1)
> ans =
>
> 0.99345 -1.24098 -1.90388 -0.98251 -0.99276
> -1.24098 7.00896 9.64769 7.12176 7.89532
> -1.90388 9.64769 14.77759 9.27637 10.70612
> -0.98251 7.12176 9.27637 8.72228 8.97276
> -0.99276 7.89532 10.70612 8.97276 11.29627
>
> ###
> octave-3.5.0+> cor(Ra)
> ans =
>
> 1.00000 -0.47029 -0.49690 -0.33377 -0.29635
> -0.47029 1.00000 0.94797 0.91085 0.88731
> -0.49690 0.94797 1.00000 0.81707 0.82863
> -0.33377 0.91085 0.81707 1.00000 0.90395
> -0.29635 0.88731 0.82863 0.90395 1.00000
>
> octave-3.5.0+> cov(Ra)
> ans =
>
> 1.0196 -1.2736 -1.9540 -1.0084 -1.0189
> -1.2736 7.1934 9.9016 7.3092 8.1031
> -1.9540 9.9016 15.1665 9.5205 10.9879
> -1.0084 7.3092 9.5205 8.9518 9.2089
> -1.0189 8.1031 10.9879 9.2089 11.5935
>
> octave-3.5.0+> cov(Ra, 1)
> ans =
>
> 0.99345 -1.24098 -1.90388 -0.98251 -0.99276
> -1.24098 7.00896 9.64769 7.12176 7.89532
> -1.90388 9.64769 14.77759 9.27637 10.70612
> -0.98251 7.12176 9.27637 8.72228 8.97276
> -0.99276 7.89532 10.70612 8.97276 11.29627
>
- mkoctfile problems with Octave 3.4.0, OSX 10.6.7, Scott Wardle, 2011/04/20
- Re: mkoctfile problems with Octave 3.4.0, OSX 10.6.7, Julien Salort, 2011/04/21
- Re: mkoctfile problems with Octave 3.4.0, OSX 10.6.7, Julien Salort, 2011/04/21
- Re: mkoctfile problems with Octave 3.4.0, OSX 10.6.7, Lukas Reichlin, 2011/04/21
- Message not available
- Problem with cor (was: Re: mkoctfile problems with Octave 3.4.0, OSX 10.6.7),
Julien Salort <=
- Re: mkoctfile problems with Octave 3.4.0, OSX 10.6.7, Scott Wardle, 2011/04/21