Re: st: tsset - repeated time values within panel

[Nick Cox <njcoxstata@gmail.com>]

There is a difference between filling in gaps and replacing missing
values. Evidently you have gaps but no missing values.

If my data run

group  year
1        2007
1        2008
1        2009
2        2007
2        2009

I have a gap for group 2 and year 2008. That is not a missing value.
You can use -tsfill- to fill in gaps but it won't add any information
to your data.

Conversely this is an example of a missing value

group year  Stata_users
2      2007    4000000
2      2008       .
2      2009    5000000

Nick

On Wed, Mar 30, 2011 at 10:38 AM, emanuele mazzini
<madsoenistata@gmail.com> wrote:

> Right. This is what I get with the command that you suggested me:
>
> egen group = group(iso1 iso2), label
>
> . tsset group year
>       panel variable:  group (unbalanced)
>        time variable:  year, 1950 to 2006, but with gaps
>                delta:  1 unit
>
> but when I say it does not work I mean that basically stata does not
> make any change in the dataset.
>
>
> 2011/3/30 Nick Cox <njcoxstata@gmail.com>:
>> I don't know what you mean by "does not work". That command should
>> work if your data have been properly -tsset-. Show us the result of
>> typing
>>
>> . tsset
>>
>> Nick
>>
>> On Wed, Mar 30, 2011 at 10:22 AM, emanuele mazzini
>> <madsoenistata@gmail.com> wrote:
>>> Hi Nick.
>>> Thank you very much for your suggestion. The command you gave me did
>>> work, but how can I replace now missing values for some given years
>>> for which I have data for the previous/following years? I tried the
>>> command: replace myvar = L.myvar if myvar >=. but it does not work.
>>>
>>> Thank you very much in advance,
>>> Emanuele.
>>>
>>> 2011/3/29 Nick Cox <njcoxstata@gmail.com>:
>>
>>>> You don't say how you are defining panels with -tsset-. It sounds as
>>>> if you regard a country pair as defining a panel, in which case
>>>> something like
>>>>
>>>> . egen group = group(cty1 cty2), label
>>>> . tsset group year
>>>>
>>>> is the way to proceed.
>>>>
>>>> On Mon, Mar 28, 2011 at 11:00 PM, emanuele mazzini
>>>> <madsoenistata@gmail.com> wrote:
>>>>
>>>>> I have a panel bilateral dataset with country pairs from year 1950 to
>>>>> 2006 and experiencing the error message repeated time values within
>>>>> panel when typing tsset. I have already read some messages sen by
>>>>> other people experiencing the same issue, and also I tried to follow
>>>>> the suggestion of the guide, but still cannot understand why i got
>>>>> that.
>>>>> To be more precise, as mine is a bilateral dataset, I have to say that
>>>>> every observation is identified by three variables (cty1, cty2 and
>>>>> year) so may be the problem relates to this, but obviously I am not
>>>>> sure.
>>>>> Finally, I am not really interested in the command per se, but the
>>>>> point is to replace some missing values which are time-constant and
>>>>> for which I do have data for the previous/following years.
>>>>> Can anybody help me to deal with this?

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/