R help - Mar 2005 - How to get standard deviation of rows in a matrix

Hi all,
 
   I am trying to find sd of my rows in a matrix and i get column sd inspite of
extracting rows.
I tried to do the sqrt(var(x)) but that did'nt work as well,
 
Here is my data
 
genes
15 24 63 40
25 42 46 35
23 53 37 45
30 37 50 55
40 51 30 48
 
x<-sd(genes[1:5,])
 
y<-sqrt(var(genes[1:5,]))
 
I get 4 sds for the 4 columns instead of 5 sds for my 5 rows.
Thanks you in advance.
 
Choudary Jagarlamudi
Instructor
Southwestern Oklahoma State University
STF 254
100 campus Drive
Weatherford OK 73096
Tel 580-774-7136

________________________________

Jagarlamudi, Choudary wrote on 3/9/2005 1:49 PM:
> Hi all,
>  
>    I am trying to find sd of my rows in a matrix and i get column sd
inspite of extracting rows.
> I tried to do the sqrt(var(x)) but that did'nt work as well,
>  
> Here is my data
>  
> genes
> 15 24 63 40
> 25 42 46 35
> 23 53 37 45
> 30 37 50 55
> 40 51 30 48
>  
> x<-sd(genes[1:5,])
>  
> y<-sqrt(var(genes[1:5,]))
>  
> I get 4 sds for the 4 columns instead of 5 sds for my 5 rows.
> Thanks you in advance.
>  
This has come up several times in the past and a quick search of the 
archives would have lead you to:

http://finzi.psych.upenn.edu/R/Rhelp02a/archive/34169.html

HTH,
--sundar

On Wed, 9 Mar 2005 13:49:44 -0600
"Jagarlamudi, Choudary" <choudary.jagar at swosu.edu> wrote:
> Hi all,
>  
>    I am trying to find sd of my rows in a matrix and i get column sd
inspite of extracting rows.
> I tried to do the sqrt(var(x)) but that did'nt work as well,
>  
> Here is my data
>  
> genes
> 15 24 63 40
> 25 42 46 35
> 23 53 37 45
> 30 37 50 55
> 40 51 30 48
>  
> x<-sd(genes[1:5,])
>  
> y<-sqrt(var(genes[1:5,]))
>  
> I get 4 sds for the 4 columns instead of 5 sds for my 5 rows.
> Thanks you in advance.
> mymat <- matrix(rnorm(20), 5)
> mymat
           [,1]       [,2]       [,3]       [,4]
[1,] -0.1418666 -0.6754704 -0.2525154 -1.4832003
[2,] -0.2254920  1.8705093 -0.9678318 -0.1108883
[3,]  2.2501392  0.1687349 -0.1279790  0.7055311
[4,]  0.9893453 -0.5924199  0.2410576  0.9001638
[5,] -0.4179559  0.9334556 -0.6501605  0.6148958
> apply(mymat, 1, sd)
[1] 0.6084171 1.2135998 1.0584750 0.7318231 0.7722641

See ?apply

HTH,
Tobias

You should read ?var and ?sd more carefully.  For a data frame or a matrix,
var() returns the covariance matrix of the columns, whereas sd() returns the
standard deviations of the columns.  If you want standard deviations of the
rows, you need to transpose the data.

Andy
> From: Jagarlamudi, Choudary
> 
> Hi all,
>  
>    I am trying to find sd of my rows in a matrix and i get 
> column sd inspite of extracting rows.
> I tried to do the sqrt(var(x)) but that did'nt work as well,
>  
> Here is my data
>  
> genes
> 15 24 63 40
> 25 42 46 35
> 23 53 37 45
> 30 37 50 55
> 40 51 30 48
>  
> x<-sd(genes[1:5,])
>  
> y<-sqrt(var(genes[1:5,]))
>  
> I get 4 sds for the 4 columns instead of 5 sds for my 5 rows.
> Thanks you in advance.
>  
> Choudary Jagarlamudi
> Instructor
> Southwestern Oklahoma State University
> STF 254
> 100 campus Drive
> Weatherford OK 73096
> Tel 580-774-7136
> 
> ________________________________
> 
>

On Wed, 9 Mar 2005 13:49:44 -0600, "Jagarlamudi, Choudary"
<choudary.jagar at swosu.edu> wrote :
>Hi all,
> 
>   I am trying to find sd of my rows in a matrix and i get column sd inspite
of extracting rows.
>I tried to do the sqrt(var(x)) but that did'nt work as well,
> 
>Here is my data
> 
>genes
>15 24 63 40
>25 42 46 35
>23 53 37 45
>30 37 50 55
>40 51 30 48
> 
>x<-sd(genes[1:5,])
This doesn't extract the rows, it just returns the same matrix.  As
the man page for sd says, when applied to a matrix it calculates the
standard deviation of each column.

What you want to do is to use the apply function.  Rows are the first
dimension, so you would use

apply(genes, 1, sd)

Duncan Murdoch