[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

From |
"Nick Cox" <[email protected]> |

To |
<[email protected]> |

Subject |
RE: st: RE: RE: feeding optimal lag lengths in levinlin andipshinwithin a loop |

Date |
Wed, 22 Oct 2003 14:59:30 +0100 |

Thanks. In fact, a closer scrutiny of your original reveals what may be the main problem, so my request for extra information should have been superfluous in this case. I link you need to call lags(A1 A2 <etc.> _not_ lags(`A1' `A2' <etc.>) I wouldn't do it that way, however: foreach k of varlist netprem disclm elr solv { preserve keep if variable == "`k'" forvalues x = 1/19 { sum lagrawnotrend if comcode == `x', mean local lags "`lags' `r(mean)'" } levinlin `k', lags(`lags') restore } I still don't buy the explanation for the -preserve- and -restore-. Sounds as if you just need to modify the -if- condition. Nick [email protected] > -----Original Message----- > From: [email protected] > [mailto:[email protected]]On Behalf Of > Dev Vencappa > Sent: 22 October 2003 14:45 > To: [email protected] > Subject: Re: st: RE: RE: feeding optimal lag lengths in levinlin > andipshinwithin a loop > > > Nick > > Sorry for missing some details in the explanation. This is > how the data looks like. Let us say in one panel I have > variables A B C D E F and suppose variable A is the optimal > lag length for each company in that panel. This one panel > however refers to an identifier variable called netprem. > Now I append to the first panel a second panel which is for > for identifier variable disclm, which takes the same values > for variables B C D E F but has different values for > variable A (as the optimal lag length will be different for > a different identifier and each company for that diferent > identifier). Same applies to the panel for identifiers elr > and solv. This is why I have to use preserve to ask stata > to run the levinlin test for one identifier, and then > restore the data to run it for the second identifier and so on. > When using this loop, the error message I get is > option lags() required > > which seems to indicate that stata cannot read these scalar > values A1, A2, etc in the levinlin option. > > Hope that helps in finding a solution to the problem. > > Many thanks > > Dev > > Dev Vencappa > School of Economics > University of Nottingham > University Park > Nottingham > NG7 2RD > U.K. > Tel : +44 (0)115 951 5608 > Fax: +44 (0) 115 951 4159 > >>> [email protected] 10/22/03 2:26 PM >>> > You don't say how this fails to work, but a lot may > hinge on > > keep if variable == "`k'" > > which suggests either an unusual data structure > or a misunderstanding of -keep-. > > However, none of the -preserve-, -restore- or > -keep- seems essential for the rest of the code. > > Also, are the results of -summarize- all integers? > > Nick > [email protected] > > > -----Original Message----- > > From: [email protected] > > [mailto:[email protected]]On Behalf Of > > Dev Vencappa > > Sent: 22 October 2003 14:15 > > To: [email protected] > > Subject: st: RE: feeding optimal lag lengths in levinlin and > > ipshinwithin a loop > > > > > > Stata users, I have a slight problem with panel unit root test. > > Thanks to an earlier suggestion by Nick, Ernest and others, > > I managed to replicate what I am doing below for simple > > time series. > > My current problem is that I have four variables recorded > > in a panel of 19 companies over 18 years. For each > > variable,I have worked out an optimal lag length for each > > company, and would now like levinlin to be able to read > > this optimal lag length for each company rather than me > > having to feed it in the lags option of the command every > > time. I have come up with the following loop, which does > > not seem to work unfortunately. Can anybody spot the > > problem please and suggest a shortcut to that if possible? > > The same thing will also be used for ipshin. Does anybody > > also know how to save the results of ipshin and call them > > into a variable? > > > > foreach k of varlist netprem disclm elr solv { > > preserve > > keep if variable=="`k'" > > forvalues x=1/19{ > > sum lagrawnotrend if comcode==`x', mean > > scalar A`x'=r(mean) > > } > > > > levinlin `k', lags(`A1' `A2' `A3' `A4' `A5' `A6' `A7' `A8' > > `A9' `A10' `A11' `A12' `A13' `A14' `A15' `A16' `A17' `A18' `A19') > > restore > > } > > > > * > * For searches and help try: > * http://www.stata.com/support/faqs/res/findit.html > * http://www.stata.com/support/statalist/faq > * http://www.ats.ucla.edu/stat/stata/ > > > * > * For searches and help try: > * http://www.stata.com/support/faqs/res/findit.html > * http://www.stata.com/support/statalist/faq > * http://www.ats.ucla.edu/stat/stata/ * * For searches and help try: * http://www.stata.com/support/faqs/res/findit.html * http://www.stata.com/support/statalist/faq * http://www.ats.ucla.edu/stat/stata/

**Follow-Ups**:

**References**:**Re: st: RE: RE: feeding optimal lag lengths in levinlin andipshinwithin a loop***From:*"Dev Vencappa" <[email protected]>

- Prev by Date:
**Re: st: Reminder time** - Next by Date:
**st: XTREGar command for a wage equation** - Previous by thread:
**Re: st: RE: RE: feeding optimal lag lengths in levinlin andipshinwithin a loop** - Next by thread:
**RE: st: RE: RE: feeding optimal lag lengths in levinlin andipshinwithin a loop** - Index(es):

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