xtfisher test

[Mario Marra]

hi all,
I have unbalanced panel data with n=27871 and T=122 (Delta= 1 Day) and I am trying to perform a xtfisher test, but I get this output:

.xtfisher price
too many values
r(134);

I cannot resolve this problem.
Someone can help me?
Thanks in advance

[Scott Merryman]

xtfisher was written in 2005 in version 8. It uses the out-of-date command levels to cycles through the panels. It you was using at least Stata version 11 than use -xtunitroot fisher-

[Mario Marra]

I tried with what you suggested me but also this does not work.
After I saw this post https://www.statalist.org/forums/for...66#post1308266
and I tried to follow what there was written in that post. I obtained this outcome:


. dfuller price if pvar==1004, lag (0)
no observations
r(2000);

Now, I download xtfisher command (ssc desc xtfisher). I ran the following command and that is the result

. xtfisher price , lag(0)
too many values
r(134);

and so I went back to the first question I asked.
what do you suggest me?

thanks in advance and sorry for my unclear English.

[Scott Merryman]

I would recommend dropping the singletons that causing a problem using -xtunitroot fisher- Something like:

bys panelvar: drop if  _N ==1

As it is currently written, you will not be able to get xtfisher to work with 27,000 panels.