Re: Testing for Positive Definiteness
- From: spellucci@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx (Peter Spellucci)
- Date: Mon, 30 Jan 2006 12:13:44 +0000 (UTC)
In article <hKtCf.19510$Kt5.5323@xxxxxxxxxxxxxxxxxxxx>,
"Jeremy Watts" <jwatts1970@xxxxxxxxxxx> writes:
>
><mhill37@xxxxxxxxxxx> wrote in message
>news:1138337286.183802.299440@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
>> Hello.
>>
>> I am wondering what is the easiest (in terms of computational effort)
>> way to determine if a matrix is positive definite. I am dealing with
>> symmetric matrices, and I know that one trait of a positive definite
>> matrix is that all its eigenvalues are positive. But rather than
>> compute all eigenvalues and then check to see if they are all greater
>> than zero, is there an easier way?
>
>Yes , carry out an LU decomposition and if the numbers along the diagonal of
w i t h o u t p i v o t i n g or with symmetric row/column interchanges !!!
hth
peter
better: -> lapack
>the upper triangular matrix are real and positive then the matrix is
>positive definite
>>
>> Regards,
>>
>>
>> Michael
>>
>
>
.
- References:
- Testing for Positive Definiteness
- From: mhill37
- Re: Testing for Positive Definiteness
- From: Jeremy Watts
- Testing for Positive Definiteness
- Prev by Date: New mathematics/physical sciences positions at http://jobs.phds.org, January 30, 2006
- Next by Date: Re: Solution of PDEs with guaranteed accuracy
- Previous by thread: Re: Testing for Positive Definiteness
- Next by thread: Does exist any equivalent of gradient method in functional spaces?
- Index(es):
Relevant Pages
|
