[BiO BB] Phylip distance matrix symmetry error

[Mike Marchywka]

Was this posted here already?

http://www.the-scientist.com/news/home/53289/


Just a reminder to check your stuff for consistency at every step:)

The problem you describe, however, should either generate a WARNING
or have an interface that doesn't request redundant values. Hacking up a 
nice
regular array can be cumbersome so it is probably easier to send duplicate
values. The code may make use of assumed relationships but you would think
the more naive brute force code would be available if constraints are
violated.




Mike Marchywka
586 Saint James Walk
Marietta GA 30067-7165
404-788-1216 (C)<- leave message
989-348-4796 (P)<- emergency only
marchywka at hotmail.com





>From: Pedro Fernandes <pfern at igc.gulbenkian.pt>
>Reply-To: "General Forum at Bioinformatics.Org" 
><bio_bulletin_board at bioinformatics.org>
>To: "General Forum at Bioinformatics.Org" 
><bio_bulletin_board at bioinformatics.org>
>Subject: Re: [BiO BB] Phylip distance matrix symmetry error
>Date: Sat, 16 Jun 2007 15:53:29 +0100
>
>Hi Samantha
>
>I may be missing the ccontext... but:
>
>1) There may be a precision problem, any real number in a computer is rich 
>in
>decimal places that are not always seen, what is displayed is less precise, 
>so:
>subtract the numbers!
>
>2) Some (!) Phylip rograms are known to be sensitive to invisible stuff on 
>input
>(insvisible being spaces, tabs, etc. So you may be evaluating somenthing 
>that is
>read with a wrong format.
>
>3) Check carefully the origin and contents of all your files. Use Notepad++ 
>for
>example.
>
>Good luck
>Pedro
>
>--
>Pedro Fernandes
>Centro Português de Bioinformática
>Instituto Gulbenkian de Ciência
>Apartado 14
>2781 OEIRAS
>PORTUGAL
>-----------------------------------------------------------
>CONFIDENCIAL. Esta mensagem e os ficheiros eventualmente anexos são
>confidenciais. Se tiver recebido esta mensagem por engano, agradecemos
>que nos contacte imediatamente por e-mail ou pelo telefone +351 21 4407912
>e que elimine a mensagem e ficheiros anexos sem os reproduzir.
>
>
>
>Quoting Samantha Fox <bioinfosm at gmail.com>:
>
> > ERROR: distance matrix is not symmetric:
> >        (42,1) element and (1,42) element are unequal.
> >        They are   1.544868 and   1.544868, respectively.
> >        Is it a distance matrix?
> >
> > That looks symmetric to me !!
> > Any clues anyone ??
> >
> > ~S
> > _______________________________________________
> > General Forum at Bioinformatics.Org - 
>BiO_Bulletin_Board at bioinformatics.org
> > https://bioinformatics.org/mailman/listinfo/bio_bulletin_board
> >
>_______________________________________________
>General Forum at Bioinformatics.Org - BiO_Bulletin_Board at bioinformatics.org
>https://bioinformatics.org/mailman/listinfo/bio_bulletin_board

_________________________________________________________________
Picture this – share your photos and you could win big!  
http://www.GETREALPhotoContest.com?ocid=TXT_TAGHM&loc=us