Dear R users,
I hope that this format is find. 
I'm a new user of R. I'm trying to do a LM test as in Terasvirta (2013) and when I run the code there is this type of error: Error in t(mX) %*% mX : requires numeric/complex matrix/vector arguments.
The problem could be on how I construct by dataset.

 I have import the data by the following command 
getwd()
[1] "C:/Program Files/R"
> list()
list()
> list.files()
[1] "colombo.csv"  "colombov.csv" "colombov.txt" "R-3.0.2"     
> D<-read.table(file="colombov.txt", header=T)
In the file colombov there are 9 variables for T=126.

> names(X)
[1] "DATE"   "G"      "T"      "Y"      "SV"     "GperSV" "TperSV" "YperSV"
[9] "news"  



>fix(D)
From my D I have to construct 3 matrices. To do that I go on with the further steps:

First of all---> I need 3 matrix(mY, mX, mZ)
I create a matrix for each of the above 9 variables. 
G<-matrix(D$G) I repeat this step for all the above varibles. (I repeat this step n=9)
Then, I use
mY<-matrix(nrow=126, ncol=4)         
mY<-cbind(G, T, Y, news)
I repeat the above step to create matrix mX and mZ. 

Then I check the class(mZ) class(mY) and class(mY).
It returns that mZ, mY, mX are matrix.

Is there something wrong when I construct my dataset? Is for this reason that I cannot compute my LM test and it returns "Error : requires numeric/complex matrix/vector arguments‏"?

Any suggestions
Thanks a lot
Valentina

Any suggestion?
Thanks a lot
Valentina