Raw Message

Message-ID: <20110429150958.GA13358@praha1.ff.cuni.cz>
Date: 2011-04-29T15:09:58Z
From: Petr Savicky
Subject: threshold matrix
In-Reply-To: <735733.79058.qm@web120104.mail.ne1.yahoo.com>

On Fri, Apr 29, 2011 at 07:44:59AM -0700, Alaios wrote:
> Thanks a lot.
> I finally used
> 
> M2 <- M
> M2[M < thresh] <- 0
> M2[M >= thresh] <- 1
> 
> as I noticed that this one line
> 
> M2 <- as.numeric( M[] < thresh )
> vectorizes my matrix.

Hi.

This may be avoided, for example

  M2 <- M
  M2[, ] <- as.numeric(M >= thresh)

or

  array(as.numeric(M >= thresh), dim=dim(M))

> One more question I have two matrices that only differ slightly. What will be the easiest way to compare and find the cells that are not the same?

If A and B are matrices of the same dimension, then

  A == B

is a logical matrix with TRUE entires for positions, where 
A and B match exactly.

  abs(A - B) <= eps

is a logical matrix with TRUE entires for positions, where
A and B differ at most by eps.

If you want to get only one logical result, then use

  all(A == B)

for exact equality and

  all(abs(A - B) <= eps)

for approximate equality of all entries.

See also ?all.equal, which uses the relative error, not absolute
difference.

Hope this helps.

Petr Savicky.