[R] Need help on upper.tri()

[Ravi Varadhan]

There are other ways to make symmetric matrices:

1.   mat + t(mat)

2.  crossprod(mat)

3.  tcrossprod(mat)

(1) is slightly faster than (2) and (3) (difference is trivial except for very large matrices), but (2) and (3) are guranteed to give you a positive-semidefinite (PSD) matrices, whereas (1) is not.

Ravi.
____________________________________________________________________

Ravi Varadhan, Ph.D.
Assistant Professor,
Division of Geriatric Medicine and Gerontology
School of Medicine
Johns Hopkins University

Ph. (410) 502-2619
email: rvaradhan at jhmi.edu


----- Original Message -----
From: William Dunlap <wdunlap at tibco.com>
Date: Tuesday, August 3, 2010 12:28 pm
Subject: Re: [R] Need help on upper.tri()
To: Nikhil Kaza <nikhil.list at gmail.com>, Ron Michael <ron_michael70 at yahoo.com>
Cc: r-help at r-project.org


> > -----Original Message-----
>  > From: r-help-bounces at r-project.org 
>  > [ On Behalf Of Nikhil Kaza
>  > Sent: Tuesday, August 03, 2010 8:56 AM
>  > To: Ron Michael
>  > Cc: r-help at r-project.org
>  > Subject: Re: [R] Need help on upper.tri()
>  > 
>  > 
>  > try using Matrix package instead
>  > 
>  > mat <- Matrix(rnorm(25),5,5)
>  > forceSymmetric(mat)
>  > 
>  > The reason your method does not work is because matrix is 
>  > effectively  
>  > a vector and the indices increase along rows within a column.
>  
>  To copy the transpose of the upper triangle to the
>  lower triangle using only base R functions try
>     mat[lower.tri(mat)] <- t(mat)[lower.tri(mat)]
>  
>  Bill Dunlap
>  Spotfire, TIBCO Software
>  wdunlap tibco.com  
>  
>  > 
>  > Nikhil
>  > 
>  > On Aug 3, 2010, at 7:36 AM, Ron Michael wrote:
>  > 
>  > > HI, I am really messing up to make a symmetrical matrix using  
>  > > upper.tri() & lower.tri() function. Here is my code:
>  > >
>  > >> set.seed(1)
>  > >> mat = matrix(rnorm(25), 5, 5)
>  > >> mat
>  > >            [,1]       [,2]       [,3]        [,4]        [,5]
>  > > [1,] -0.6264538 -0.8204684  1.5117812 -0.04493361  0.91897737
>  > > [2,]  0.1836433  0.4874291  0.3898432 -0.01619026  0.78213630
>  > > [3,] -0.8356286  0.7383247 -0.6212406  0.94383621  0.07456498
>  > > [4,]  1.5952808  0.5757814 -2.2146999  0.82122120 -1.98935170
>  > > [5,]  0.3295078 -0.3053884  1.1249309  0.59390132  0.61982575
>  > >> mat[lower.tri(mat)] = mat[upper.tri(mat)]
>  > >> mat
>  > >             [,1]        [,2]        [,3]        [,4]        [,5]
>  > > [1,] -0.62645381 -0.82046838  1.51178117 -0.04493361  0.91897737
>  > > [2,] -0.82046838  0.48742905  0.38984324 -0.01619026  0.78213630
>  > > [3,]  1.51178117 -0.01619026 -0.62124058  0.94383621  0.07456498
>  > > [4,]  0.38984324  0.94383621  0.78213630  0.82122120 -1.98935170
>  > > [5,] -0.04493361  0.91897737  0.07456498 -1.98935170  0.61982575
>  > >
>  > >
>  > > Which is not coming as symmetrical function. Can anyone 
>  > point me on  
>  > > the correct way of using upper, lower.try() function to get a  
>  > > symmetrical matrix?
>  > >
>  > > Thanks,
>  > >
>  > >
>  > > 	[[alternative HTML version deleted]]
>  > >
>  > > ______________________________________________
>  > > R-help at r-project.org mailing list
>  > > 
>  > > PLEASE do read the posting guide 
>  > 
>  > > and provide commented, minimal, self-contained, reproducible code.
>  > 
>  > ______________________________________________
>  > R-help at r-project.org mailing list
>  > 
>  > PLEASE do read the posting guide 
>  > 
>  > and provide commented, minimal, self-contained, reproducible code.
>  > 
>  
>  ______________________________________________
>  R-help at r-project.org mailing list
>  
>  PLEASE do read the posting guide 
>  and provide commented, minimal, self-contained, reproducible code.