Hi r lovers!
I 'd like to apply the cor() function to a matrix which have some missing
values
As a matter of fact and quite logically indeed it doesn't work
Is there a trick to replace the missing value by the mean of each variable or by
any other relevant figures ?
Or should I apply a special derivate of the cor() function, (I don't have
any idea if it exists and have some trouble to figure out how it could)
to skip this trouble?
Thanks a lot for any suggestions and help
Vincent
*************************************************************************
Ce message et toutes les pieces jointes (ci-apres le "message") sont
confidentiels et etablis a l'intention exclusive de ses destinataires.
Toute utilisation ou diffusion non autorisee est interdite.
Tout message electronique est susceptible d'alteration.
La SOCIETE GENERALE et ses filiales declinent toute responsabilite au
titre de ce message s'il a ete altere, deforme ou falsifie.
********
This message and any attachments (the "message") are confidentia...
{{dropped}}
On Thu, 24 Apr 2003 12:27:50 +0200 vincent.stoliaroff at socgen.com wrote:> > > Hi r lovers! > > I 'd like to apply the cor() function to a matrix which have some missing values > As a matter of fact and quite logically indeed it doesn't work > Is there a trick to replace the missing value by the mean of each variable or by any other relevant figures ? > Or should I apply a special derivate of the cor() function, (I don't have any idea if it exists and have some trouble to figure out how it could) > to skip this trouble? > Thanks a lot for any suggestions and help > > Vincent > >Even though using pairwise deletion of NAs will result sometimes in a singular correlation matrix, it is better to do that than to replace NAs with constants, which will distort the correlations. You may want to look at the rcorr function in the Hmisc package, which does pairwise deletion of NAs for Pearson and Spearman correlations. See http://hesweb1.med.virginia.edu/biostat/s/Hmisc.html> > > > ************************************************************************* > Ce message et toutes les pieces jointes (ci-apres le "message") sont > confidentiels et etablis a l'intention exclusive de ses destinataires. > Toute utilisation ou diffusion non autorisee est interdite. > Tout message electronique est susceptible d'alteration. > La SOCIETE GENERALE et ses filiales declinent toute responsabilite au > titre de ce message s'il a ete altere, deforme ou falsifie. > ******** > This message and any attachments (the "message") are confidentia... {{dropped}}Please do not include such long disclaimers in your messages. --- Frank E Harrell Jr Prof. of Biostatistics & Statistics Div. of Biostatistics & Epidem. Dept. of Health Evaluation Sciences U. Virginia School of Medicine http://hesweb1.med.virginia.edu/biostat
replace.na.m <- function (x){
X<-mean(x,na.rm=TRUE)
ifelse ( is.na(x)=="TRUE",X,x)
}
apply(data.frame,2,replace.na.m)
another way is using cor() and the option
use: an optional character string giving a method for computing
covariances in the presence of missing values. This must be
(an abbreviation of) one of the strings `"all.obs"',
`"complete.obs"' or
`"pairwise.complete.obs"'.
perhaps this helps you,
regards,christian
----- Original Message -----
From: <vincent.stoliaroff at socgen.com>
To: <r-help at stat.math.ethz.ch>
Sent: Thursday, April 24, 2003 12:27 PM
Subject: [R] Missing Value And cor() function
>
>
> Hi r lovers!
>
> I 'd like to apply the cor() function to a matrix which have some
missing
values> As a matter of fact and quite logically indeed it doesn't work
> Is there a trick to replace the missing value by the mean of each variable
or by any other relevant figures ?> Or should I apply a special derivate of the cor() function, (I don't
have
any idea if it exists and have some trouble to figure out how it
could)> to skip this trouble?
> Thanks a lot for any suggestions and help
>
> Vincent
>
>
>
>
>
> *************************************************************************
> Ce message et toutes les pieces jointes (ci-apres le "message")
sont
> confidentiels et etablis a l'intention exclusive de ses destinataires.
> Toute utilisation ou diffusion non autorisee est interdite.
> Tout message electronique est susceptible d'alteration.
> La SOCIETE GENERALE et ses filiales declinent toute responsabilite au
> titre de ce message s'il a ete altere, deforme ou falsifie.
> ********
> This message and any attachments (the "message") are
confidentia...
{{dropped}}>
> ______________________________________________
> R-help at stat.math.ethz.ch mailing list
> https://www.stat.math.ethz.ch/mailman/listinfo/r-help
On Thu, 24 Apr 2003 vincent.stoliaroff at socgen.com wrote:> > > Hi r lovers! > > I 'd like to apply the cor() function to a matrix which have some > missing values As a matter of fact and quite logically indeed it doesn't > work Is there a trick to replace the missing value by the mean of each > variable or by any other relevant figures ? Or should I apply a special > derivate of the cor() function, (I don't have any idea if it exists and > have some trouble to figure out how it could) to skip this trouble?The cor() function has two options for handling missing values, which are described on its help page. -thomas