[R] vardiag package help

Monica Palaseanu-Lovejoy monica.palaseanu-lovejoy at stud.man.ac.uk
Fri Aug 29 15:49:01 CEST 2003


Hi,

First of all a big THANK YOU to all who answered me 
yesterday. 

I am back to my problems with outliers. I did a qqnorm on my 
data (as somebody suggested), but I will like to compare results 
doing other plots as well. I used ‘cook.distance’ on lm and glm 
objects, but still I am not happy with that. 

So now I am loading the “vardiag” package. First thing is to 
transform my data (frame.data) into a matrix – which I did with 
‘as.matrix’. I verify my new object as being amatrix – and it is, 
and I list it to be sure that in is a matrix of n rows with 3 colums.

Now I want to use the command ‘varobj’ to transform my 
matrix into a variogram object.

The param are:
varobj(m, iter = 50, tolerance = 2e-04, trace = 1, loo = F)
where, m is a n by 3 matrix with spatial data.

My matrix has a column for x value, for y value and for pah16 
concentrations.

I always get the following error:
Error in matrix(i1[row(i1) < col(i1)], n1, n1) : 
        cannot allocate vector of length 1223600400

What does this means and what I am doing wrong?

Thank you in advance,

Monica

Monica Palaseanu-Lovejoy
University of Manchester
School of Geography
Mansfield Cooper Building 
Oxford Road, Manchester
M13 9PL, UK. 
email: monica.palaseanu-lovejoy at stud.man.ac.uk




More information about the R-help mailing list