Even if I don't include interaction terms (i.e.,
x1+x2+x3) it gives me the error, however, if I remove the binary variables it works fine.
Advice???
S
On Tue, Sep 13, 2011 at 8:48 AM, Kosuke Imai <kimai(a)princeton.edu> wrote:
I don't think x1*(x2+x3) is a valid syntax (it may be but may not be accepted by
Zelig). If you want interaction term, you need to do x1:x2 or x1*x2 if you want to also
include main effects.
Kosuke
Department of Politics
Princeton University
Hi!
I am trying to calculate ATT for some data, and I am getting an error. Below is the
relevant code, the error and the traceback...
znn.out<-zelig(growthz~age+cogat+iep*(mapfall+itbs)+lunch*(mapfall+itbs)+treat,
data=match.data(nearest), model="ls")
nnx.out <- setx(znn.out, data = match.data(nearest,"treat"), cond = TRUE)
snn.out <- sim(znn.out, x = nnx.out)
Error in mvrnorm(num, mu = coef(object), Sigma = vcov(object)) :
incompatible arguments
traceback()
6: stop("incompatible arguments")
5: mvrnorm(num, mu = coef(object), Sigma = vcov(object))
4: param.lm(object, num = num, bootstrap = bootstrap)
3: param(object, num = num, bootstrap = bootstrap)
2: sim.cond(znn.out, x = nnx.out)
1: sim(znn.out, x = nnx.out)
Note: Age, cogat, mapfall, and itbs are continuous variables. Iep, lunch, and treat are
binary.
Please help!
S