Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: RE: St: No observations error message

From   [email protected]
To   [email protected]
Subject   Re: st: RE: St: No observations error message
Date   Thu, 12 Aug 2004 17:01:42 -0500

Nick and Kit, 

Following both of your suggestions, I double checked that I have -tsset and ran
xtdes.  Everything looks normal-- it gives me my -n and -T dimensions
correctly.  So I -tsreport with all options, and it reported that I have no
gaps in sample.  Everything else looks good-- all strings are described as
string, and numbers as numbers.

So, in the meantime, I don't have a working dataset.  If anything else comes to
mind, please let me know.


>Are you sure that you have tsset by both individual and time 
indicators, so that tsset reports something like

. tsset
        panel variable:  company, 1 to 10
         time variable:  year, 1935 to 1954

Try running xtdes and see what happens.

Gaps are not allowed in most of Stata's time series commands. That is, 
you need a time series that is complete. In the above example, a 
missing value in 1950 would be fatal. It may be that you have gaps in 
that the data are not properly tsset, and Stata is taking the 
difference between company 1, year 1954 and company 2, 1935 as a gap. 
If data are tsset properly, "panel gaps" will be ignored by commands 
that cannot tolerate gaps, such as those you have described above.

KitQuoting Nick Cox <[email protected]>:

> [email protected]
> > I am trying to run -xtreg, fe.  When I do this, stata 
> > claims:-r(2000) no
> > observations.  I also get this when I try _varlag.  What 
> > could be the issue
> > here? I have the data -tsset, and data is certainly in the 
> > data set.  Any other
> > troubleshooting tips?
> Possibly one of the variables in your list is actually 
> string even though it should be numeric. Or the pattern
> of missing values means that you have no observations 
> for which all the variables are not missing. 
> > Additionally, in trying to run any panel unit root tests or 
> > -johans, I am told
> > my data "has gaps".  What is this and how might I correct it?
> Your data possess gaps. You need to fill them in or live
> with not running these tests. Read up about -tsreport-. 
> Nick 
> [email protected] 
> *
> *   For searches and help try:
> *
> *
> *

*   For searches and help try:

© Copyright 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index