Hi,
I'm trying to fit a fixed effect (LSDV) panelmodel with R. I have a dataset
with y as dependent, x1&x2 as indeps, t as time index and i as an
id-variable for each individual. There are three observations for each
individual (t=1, t=2, t=3).
I want to try a simple regression, but with individual intercepts:
-------------------------------------------------
# reading in some data ...
mydata <- read.csv(...)
attach(mydata)
# fit modell
mymodel <- lm(y ~ -1 + factor(i) + x1 + x2)
summary(mymodel)
-------------------------------------------------
Works fine when the size of my dataset doesn't exceed about n=5000
observations, but I have some more. Can I do a partitioned regression with
R, are there any other options already implemented in R ?
Thanks,
Thomas