[David Achiles]

Dear Nick

you are right I tried

cycleplot (SSC) doesn't apply directly to quarterly data.

and it is not working as well

[Nick Cox]

Sorry, I didn't mean what I said. I meant that cycleplot doesn't apply directly to panel data.

There is absolutely no difficulty with (single-panel) quarterly data so long as you have appropriate variables. For example, this works. It's not interesting about seasonality, but it's proof of concept.

webuse wpi1
gen quarter = quarter(dofq(t))
gen year = year(dofq(t))
cycleplot ln_wpi quarter year , length(4)

I can't comment on the code which you report as "not working" as you don't tell us what it is. But I will guess that you tried to feed it inappropriate time variables.

[David Achiles]

Nick, yes I am mean in my data it is not working as I run

cycleplot lnp1  quarter year3 , length(5)

I got this weird figure

[Nick Cox]

I guess that is a consequence of the limited length of your data and your applying it to a panel dataset. I didn't pick up from your early posts that you have such very short panels.

[David Achiles]

right I have very short panel data

by the way I attached my data in my previous post

http://www.statalist.org/forums/forum/general-stata-discussion/general/1331411-how-to-make-graph-for-seasonality-data?p=1331550#post1331550

Please could you look to data and if possible give to me some idea ?

Thank you

[Carole J. Wilson]

Post your data using the program dataex. Type ssc install dataex and read the help file: help dataex.