comp.lang.idl-pvwave archive: archive » Re: failed matrix inversion returns input-- interesting

Home » Public Forums » archive » Re: failed matrix inversion returns input-- interesting

Show: Today's Messages :: Show Polls :: Message Navigator
E-mail to friend

Re: failed matrix inversion returns input-- interesting [message #49305 is a reply to message #49302]

Fri, 14 July 2006 00:30

mmeron
Messages: 44
Registered: October 2003

Member

In article <1152860961.931059.317520@35g2000cwc.googlegroups.com>, "m_schellens@hotmail.com" <m_schellens@hotmail.com> writes:
>
> Ed Hyer wrote:
>> Can someone explain this behavior?
>> IDL> testm=[[0,0,0,0],[0,1,0,0],[0,0,1,0],[0,0,0,1]]
>> IDL> testmi=invert(testm,status)
>> IDL> print,status; 0=success, 1=fail, 2=pivot used
>> 1
>> IDL> print,testmi
>> 0.00000 0.00000 0.00000 0.00000
>> 0.00000 1.00000 0.00000 0.00000
>> 0.00000 0.00000 1.00000 0.00000
>> 0.00000 0.00000 0.00000 1.00000
>>
>> Is there any reason in creation why IDL simply copies the input into
>> the output if it cannot do the inversion? That seems, well, malevolent,
>> unless I'm missing something.
>
> Well, what do you suggest is better?
>
Well, nearly anything would be better.

I've a routine, called SVD_INVERT, which returns a regular inverse
when one exists, else it returns and inverse in the "SVD sense".
Would the IDL routine have been doing this, that would be fine (though
it should've been documented). But, it is not doing this either.
So IMO, it would've been preferable to return an array of NANs in such
case.

Mati Meron | "When you argue with a fool,
meron@cars.uchicago.edu | chances are he is doing just the same"

Report message to a moderator

[Message index]

		Re: failed matrix inversion returns input-- interesting By: news.verizon.net on Fri, 14 July 2006 05:23
		Re: failed matrix inversion returns input-- interesting By: mmeron on Fri, 14 July 2006 00:30
		Re: failed matrix inversion returns input-- interesting By: Bringfried Stecklum on Fri, 14 July 2006 00:14
		Re: failed matrix inversion returns input-- interesting By: marc schellens[1] on Fri, 14 July 2006 00:09

Previous Topic:	Re: IDL 6.3 for Mac OS X on Intel Now Available
Next Topic:	Re: IDL and Macs. Speed is not only about squared roots

-=] Back to Top [=-

[ Syndicate this forum (XML) ] [

] [

]

Current Time: Sun Oct 12 16:33:41 PDT 2025

Total time taken to generate the page: 0.00193 seconds