Hello guys!
Sorry to bother with such a question
I was trying to generate a monte carlo simulation with heteroskedasticity
errors. but I am not sure if the command line that I had
wrote is quite correct.
the type of heteroskedasticity that I want to create is such as var(e) =
var(x^4)
I began my work with this
x<- rnorm (100, 2,0.4) # generating an indepedent random variable
e<- rnorm(100,0,x^2) # generating the error term
y.fitted<- 0.5*(x) +3
y<- y.fitted+e
And then I wrote an for loop code which could give me the results of a
t-test.
It seems that the simulation worked good because the main result is that the
t-test was no longer meanfull (as it is expected)
But my main doubt is if this code was capable to generate the type of
heteroskedasticity
that I wanted.
I tried to understand how the rnorm function generates the numbers but i
could not grab much....
So could anyone help me telling if this kind of work is okay? or is there
any moree inteligent way to do the job?
thanks!!!
[[alternative HTML version deleted]]
______________________________________________
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
