Hi,
I'm trying to calculate the residuals in a one-sided AR(p) model:
sum (phi_j (X_t-j - EX)) = epsilon
My X is an ARMA(1,1)-model, which can be represented as an AR(infty) model.
I calculate the order of the model with AIC and the parameters with the
Yule-Walker method.
For the residuals, I
Good day everyone,
I have about 500 values to generate my ecdf.
Now I want to sample 2000 values i.i.d. distributed according to my ecdf.
Is it possible to sample from an ecdf?
If yes, I would appreciate your help.
Martin
--
View this message in context:
Say F_n is my ecdf, generated from my sample x.
With sample(x, 2000, replace = TRUE) I get a uniformly distributed sample
from x.
But I need a (re)sample, which is F_n distributed.
Now I'm confused, if that makes a difference or not...
Regards,
Martin
Dimitris Rizopoulos-4 wrote:
say 'x'
Hi,
I want to sample a block of information.
Let's say x is a time series. Using sample(x,5) I get a random sample of
length 5 from x.
Is it possible to sample consecutive observations, i.e. I sample one
observation and also get the next 4 observations?
Thanks a lot!
Martin
--
View this
4 matches
Mail list logo