Re: st: sroot user written pgm

[Nick Cox <njcoxstata@gmail.com>]

"Son, you're on your own." (Reverend Johnson in Blazing Saddles)

More seriously, no thoughts except don't do that. If you test each
panel separately, there is a real problem of multiplicity. If you work
with a fixed level, then bear in mind that implies so many rejections
even if all the nulls are correct and the tests. Also, are the tests
really independent of each other?
Nick
njcoxstata@gmail.com


On 4 November 2013 21:08, Francis, Richard N <rnfrancis@utep.edu> wrote:
> Nick,
>
> Most grateful for your insight, your suggestions work beautifully!
>
> Any thoughts on obtaining the test statistics from the user written pgm for further analysis (such as separating significant and insignificant results)?
>
> Again, I am grateful for the assistance available from you and the Statalist in general.
>
> Rick Francis
> Associate Professor
> Department of Accounting
> College of Business Administration
> University of Texas at El Paso
> 500 W. University Avenue
> El Paso, TX  79968
> Office: 915-747-7953
> FAX: 915-747-8618
> Email:  rnfrancis@utep.edu
>
> -----Original Message-----
> From: owner-statalist@hsphsun2.harvard.edu [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of Nick Cox
> Sent: Monday, November 04, 2013 5:52 AM
> To: statalist@hsphsun2.harvard.edu
> Subject: Re: st: sroot user written pgm
>
> Second -preserve- should be -restore-.
> Nick
> njcoxstata@gmail.com
>
>
> On 4 November 2013 10:57, Nick Cox <njcoxstata@gmail.com> wrote:
>> You are asked in the Statalist FAQ (a) to explain where user-written
>> programs you refer to come from (b) not to send attachments to the
>> list.
>>
>> With your data, -tsset gvkey obsno, quarterly- fails because -gvkey- is string.
>>
>> You would be better using a proper quarterly date:
>>
>> gen date = yq(year, fqtr)
>> tsset id date
>>
>> But your main problem appears to be that -sroot- (SJ) does not support
>> -if- correctly.  You should contact the author.
>>
>> A work around is this:
>>
>> egen id = group(gvkey), label
>> gen date = yq(year, fqtr)
>> tsset id date
>>
>> su id, meanonly
>>
>> forval i = 1/`r(max)' {
>>              preserve
>>              keep if id == `i'
>>              sroot fcf
>>              preserve
>> }
>>
>> Nick
>> njcoxstata@gmail.com
>>
>>
>> On 4 November 2013 03:28, Francis, Richard N <rnfrancis@utep.edu> wrote:
>>
>>> Anyone know if a user written program such as "sroot" (seasonal unit root test) is capable of looping such that the program tests the time series for each subject in the sample?
>>>
>>> I attached a sample with three subjects, 60 obs each.
>>>
>>> Then used the following pgm:
>>>
>>> tsset gvkey obsno, quarterly
>>>
>>> egen id = group(gvkey)
>>> su id, meanonly
>>>
>>> foreach i of num 1/`r(max)' {
>>> sroot fcf if id == `i'
>>> }
>>>
>>> Program runs, but pools the subjects rather than test them individually.
>>>
>>> I am grateful for any ideas here.
> *
> *   For searches and help try:
> *   http://www.stata.com/help.cgi?search
> *   http://www.stata.com/support/faqs/resources/statalist-faq/
> *   http://www.ats.ucla.edu/stat/stata/
>
> *
> *   For searches and help try:
> *   http://www.stata.com/help.cgi?search
> *   http://www.stata.com/support/faqs/resources/statalist-faq/
> *   http://www.ats.ucla.edu/stat/stata/
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/