Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.

# Re: st: Confirming whether a variable is binary or continuous

 From Bert Jung To statalist@hsphsun2.harvard.edu Subject Re: st: Confirming whether a variable is binary or continuous Date Mon, 26 Mar 2012 16:18:48 -0400

```Thanks all.  I wanted to apologize for my silence on this thread as I
was recovering from a bout of flu.  Nick is indeed right and my goal
is data management.  And Daniel is right that my coding was
convoluted.  I will ponder more, and read up on Cameron's suggested
papers for a related problem I see coming down the road.

Bert

On Mon, Mar 19, 2012 at 3:34 PM, Nick Cox <njcoxstata@gmail.com> wrote:
> I flagged the importance of remembering missings earlier today.
>
> In the same spirit -findname- (SJ, SSC)
>
> findname , type(numeric) all(@ == int(@))
>
> finds all integer-valued variables.
>
> Nick
>
> On Mon, Mar 19, 2012 at 7:24 PM, Seed, Paul <paul.seed@kcl.ac.uk> wrote:
>> It may be that a binary variable contains missing values.
>> For system missing, with 3 possible values [0, 1, .] one can use
>>        assert var^2 == var
>> But this still leaves the possibility of no ones or zeros,
>> and does not handle other non-missing values.
>>
>> A more general solution for any 2 non-missing values is
>>
>>        su var, mean
>>        assert (var == r(min) | var == r(max) | var >= . ) & r(min) != r(max)
>>
>> the possibility of categorical must also be considered.
>> For this there is a simple test; provided your continuous variable
>> is not rounded to the nearest integer & your categorical variables
>> do not include fractions.
>>        assert var == int(var)
>>
>>
>> Paul T Seed, Senior Lecturer in Medical Statistics,
>> Division of Women’s Health, King’s College London
>> Women’s Health Academic Centre, KHP
>> (+44) (0) 20 7188 3642.
>>
>>
>>
>>
>>
>>> Date: Sat, 17 Mar 2012 07:05:00 +0000
>>> From: Nick Cox <njcoxstata@gmail.com>
>>> Subject: Re: st: Confirming whether a variable is binary or continuous
>>>
>>> Variations on this are often used inside Stata commands
>>>
>>> sort var
>>> assert var == var[1] | var == var[_N]
>>>
>>> Here -var- must be one of two distinct values, but nothing specifies
>>> what they are precisely.
>>>
>>> Nick
>>>
>>> On Fri, Mar 16, 2012 at 11:39 PM, daniel klein
>>> > Nick,
>>> >
>>> > very smart solution, thanks. Seems it is often the simple solutions
>>> > that are so hard to find.
>>> >
>>> > The difference to the "-tabulate- solution" is that <var> is allowed
>>> > to be a constant here, which is perfectly alright for a binary
>>> > variable. It is for Bert to decide whether the indicators are allowed
>>> > to be constant or have to vary.
>>> >
>>> > Daniel
>>> >
>>> > --
>>> > You can also do this by e.g.
>>> >
>>> > assert inlist(var, 0. 1)
>>> *
>>> *   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/
>>>
>>> ------------------------------
>>>
>>> Date: Sat, 17 Mar 2012 07:46:20 +0000
>>> From: Nick Cox <njcoxstata@gmail.com>
>>> Subject: Re: st: Confirming whether a variable is binary or continuous
>>>
>>> My wording was sloppy. That test is satisfied also by one distinct
>>> value. For there to be precisely two distinct values, it is also
>>> necessary for -var[1]- to differ from -var[_N]-.
>>>
>>> Similarly the test -inlist(var, 0, 1)- is satisfied if -var- is always
>>> 0 or if -var- is always 1.
>>>
>>> Nick
>>>
>>> On Sat, Mar 17, 2012 at 7:05 AM, Nick Cox <njcoxstata@gmail.com> wrote:
>>> > Variations on this are often used inside Stata commands
>>> >
>>> > sort var
>>> > assert var == var[1] | var == var[_N]
>>> >
>>> > Here -var- must be one of two distinct values, but nothing specifies
>>> > what they are precisely.
>>> >
>>> > Nick
>>> >
>>> > On Fri, Mar 16, 2012 at 11:39 PM, daniel klein
>>> >> Nick,
>>> >>
>>> >> very smart solution, thanks. Seems it is often the simple solutions
>>> >> that are so hard to find.
>>> >>
>>> >> The difference to the "-tabulate- solution" is that <var> is allowed
>>> >> to be a constant here, which is perfectly alright for a binary
>>> >> variable. It is for Bert to decide whether the indicators are
>>> allowed
>>> >> to be constant or have to vary.
>>> >>
>>> >> Daniel
>>> >>
>>> >> --
>>> >> You can also do this by e.g.
>>> >>
>>> >> assert inlist(var, 0. 1)
>>> *
>>> *   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/
>>>
>>> ------------------------------
>>>
>>> Date: Sat, 17 Mar 2012 09:03:04 +0000 (GMT)
>>> From: olorunfemi sola <solafem7@yahoo.co.uk>
>>> Subject: st: working with panel data in Stata
>>>
>>> Dear All,
>>>
>>> Am working with panel data. The data are as follows:
>>> id    time    ep    dep    lr
>>> lagos    2000    10    11    12
>>> lagos    2001    8    12    13
>>> lagos    2002    9    9    9
>>> lagos    2003    7    10    12
>>> ondo    2000    10    8    10
>>> ondo    2001    8    12    13
>>> ondo    2002    10    11    14
>>> ondo    2003    7    10    11
>>> ife    2000    9    11    12
>>> ife    2001    8    8    11
>>> ife    2002    9    10    12
>>> ife    2003    10    9    12
>>> But when i give this command --------" tsset id time " it gave the
>>> following error:
>>>   tsset id time
>>> varlist:  id:  string variable not allowed
>>> r(109)
>>>
>>>
>>> ***********************************************************************
>>> SOLA OLORUNFEMI   Ph.D
>>> SENIOR LECTURER
>>>  DEAPARTMENT OF ECONOMICS
>>> AKUNGBA AKOKO
>>> ONDO STATE NIGERIA
>>> TEL NO +234 803 581 0893
>>>
>>> **********************************************************************
>>>
>>>
>>> ________________________________
>>> From: Nick Cox <njcoxstata@gmail.com>
>>> To: statalist@hsphsun2.harvard.edu
>>> Sent: Friday, 16 March 2012, 23:46
>>> Subject: Re: st: Confirming whether a variable is binary or continuous
>>>
>>> My wording was sloppy. That test is satisfied also by one distinct
>>> value. For there to be precisely two distinct values, it is also
>>> necessary for -var[1]- to differ from -var[_N]-.
>>>
>>> Similarly the test -inlist(var, 0, 1)- is satisfied if -var- is always
>>> 0 or if -var- is always 1.
>>>
>>> Nick
>>>
>>> On Sat, Mar 17, 2012 at 7:05 AM, Nick Cox <njcoxstata@gmail.com> wrote:
>>> > Variations on this are often used inside Stata commands
>>> >
>>> > sort var
>>> > assert var == var[1] | var == var[_N]
>>> >
>>> > Here -var- must be one of two distinct values, but nothing specifies
>>> > what they are precisely.
>>> >
>>> > Nick
>>> >
>>> > On Fri, Mar 16, 2012 at 11:39 PM, daniel klein
>>> >> Nick,
>>> >>
>>> >> very smart solution, thanks. Seems it is often the simple solutions
>>> >> that are so hard to find.
>>> >>
>>> >> The difference to the "-tabulate- solution" is that <var> is allowed
>>> >> to be a constant here, which is perfectly alright for a binary
>>> >> variable. It is for Bert to decide whether the indicators are
>>> allowed
>>> >> to be constant or have to vary.
>>> >>
>>> >> Daniel
>>> >>
>>> >> --
>>> >> You can also do this by e.g.
>>> >>
>>> >> assert inlist(var, 0. 1)
>>> *
>>> *   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/
>>>
>>> *
>>> *   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/
>>>
>>> ------------------------------
>>>
>>> Date: Sat, 17 Mar 2012 18:26:17 +0900
>>> From: "Joseph Coveney" <jcoveney@bigplanet.com>
>>> Subject: st: Re: working with panel data in Stata
>>>
>>> Try:
>>>
>>> encode id, generate(ID) label(IDs)
>>> tsset ID time, yearly
>>>
>>> Joseph Coveney
>>>
>>> P.S.  You might want to re-read Section 2.3 ("How do I send questions
>>> to
>>> Statalist?") of the list's FAQ (URL below).
>>>
>>> - -----Original Message-----
>>>
>>> Am working with panel data. The data are as follows:
>>> id    time    ep    dep    lr
>>> lagos    2000    10    11    12
>>> lagos    2001    8    12    13
>>> lagos    2002    9    9    9
>>> lagos    2003    7    10    12
>>> ondo    2000    10    8    10
>>> ondo    2001    8    12    13
>>> ondo    2002    10    11    14
>>> ondo    2003    7    10    11
>>> ife    2000    9    11    12
>>> ife    2001    8    8    11
>>> ife    2002    9    10    12
>>> ife    2003    10    9    12
>>> But when i give this command --------" tsset id time " it gave the
>>> following
>>> error:
>>>   tsset id time
>>> varlist:  id:  string variable not allowed
>>> r(109)
>>>
>>>
>>> *
>>> *   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/
>>>
>>> ------------------------------
>>>
>>> Date: Sat, 17 Mar 2012 18:32:24 +0900
>>> From: "Joseph Coveney" <jcoveney@bigplanet.com>
>>> Subject: st: Re: Re: working with panel data in Stata
>>>
>>> Make that Section 2.2.  Sorry.
>>>
>>> Joseph Coveney
>>>
>>>
>>>
>>> *
>>> *   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/
>>>
>>> ------------------------------
>>>
>>> Date: Sat, 17 Mar 2012 09:46:04 +0000
>>> From: Nick Cox <njcoxstata@gmail.com>
>>> Subject: Re: st: Re: Re: working with panel data in Stata
>>>
>>> I think Joseph has in mind
>>>
>>> Again, even if you delete the previous posting’s contents and change
>>> its title, such practice also messes up archiving."
>>>
>>>
>>> On Sat, Mar 17, 2012 at 9:32 AM, Joseph Coveney
>>> <jcoveney@bigplanet.com> wrote:
>>> > Make that Section 2.2.  Sorry.
>>> >
>>> > Joseph Coveney
>>> >
>>> >
>>> >
>>> > *
>>> > *   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/
>>>
>>> *
>>> *   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/
>>>
>>> ------------------------------
>>>
>>> Date: Sat, 17 Mar 2012 07:23:17 -0230
>>> From: Matthew Kerby <kerbym@mun.ca>
>>> Subject: st: Parsing words on allcaps
>>>
>>> Dear Statalist
>>>
>>> Perhaps somebody can advise on how to solve this problem.
>>>
>>> I have a dataset which consists of a single string variable. The each
>>> observation consists of a sentence which in turn contains a name in all
>>> caps. I would like to create a new variable which contains only those
>>> words which appear in all caps. eg. in example below I would like to
>>> extract the name of the hockey player.
>>>
>>>
>>> clear
>>> input str244 string
>>> "GRETZKY W. likes hamburgers"
>>> "Hotdogs are preferred by LEMIEUX M."
>>> "All things being equal, CROSBY S. will eat doughnuts"
>>> end
>>>
>>> I would like to end up with a string variable (player) which looks like
>>> the following:
>>>
>>> player
>>> "GRETZKY W."
>>> "LEMIEUX M."
>>> "CROSBY S."
>>>
>>> Any help on how to parse a string variable on allcaps is greatly
>>> appreciated.
>>> Cheers,
>>> M.
>>>
>>>
>>>
>>> - --
>>> Matthew Kerby, PhD
>>>
>>> Assistant Professor
>>> Department of Political Science
>>> Memorial University of Newfoundland
>>> St. John's, Newfoundland, A1B 3X9, Canada
>>>
>>> email: kerbym(at)mun.ca
>>> web: matthewkerby.net
>>> Tel: (709) 864-3093
>>> Fax: (709) 864-4000
>>>
>>>
>>> This electronic communication is governed by the terms and conditions
>>> at
>>> http://www.mun.ca/cc/policies/electronic_communications_disclaimer_2011
>>> .php
>>>
>>> *
>>> *   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/
>>>
>>> ------------------------------
>>>
>>> Date: Sat, 17 Mar 2012 09:55:33 +0000 (GMT)
>>> From: olorunfemi sola <solafem7@yahoo.co.uk>
>>> Subject: Re: st: Re: Re: working with panel data in Stata
>>>
>>> Nick,
>>> Â Sorry, it was a costly mistake from me. I will not repeat it. Am
>>> sorry.
>>>
>>> Â
>>> ***********************************************************************
>>> SOLA OLORUNFEMIÂ Â  Ph.D
>>> SENIOR LECTURERÂ Â
>>> Â DEAPARTMENT OF ECONOMICS
>>> AKUNGBA AKOKO
>>> ONDO STATE NIGERIA
>>> TEL NO +234 803 581 0893
>>> **********************************************************************
>>>
>>>
>>> - ----- Original Message -----
>>> From: Nick Cox <njcoxstata@gmail.com>
>>> To: statalist@hsphsun2.harvard.edu
>>> Cc:
>>> Sent: Saturday, 17 March 2012, 1:46
>>> Subject: Re: st: Re: Re: working with panel data in Stata
>>>
>>> I think Joseph has in mind
>>>
>>> Again, even if you delete the previous postingâ€™s contents and change
>>> its title, such practice also messes up archiving."
>>>
>>>
>>> On Sat, Mar 17, 2012 at 9:32 AM, Joseph Coveney
>>> <jcoveney@bigplanet.com> wrote:
>>> > Make that Section 2.2. Â Sorry.
>>> >
>>> > Joseph Coveney
>>> >
>>> >
>>> >
>>> > *
>>> > * Â  For searches and help try:
>>> > * Â  http://www.stata.com/help.cgi?search
>>> > * Â  http://www.stata.com/support/statalist/faq
>>> > * Â  http://www.ats.ucla.edu/stat/stata/
>>>
>>> *
>>> *Â   For searches and help try:
>>> *Â  http://www.stata.com/help.cgi?search
>>> *Â  http://www.stata.com/support/statalist/faq
>>> *Â  http://www.ats.ucla.edu/stat/stata/
>>>
>>>
>>> *
>>> *   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/
>>>
>>> ------------------------------
>>>
>>> Date: Sat, 17 Mar 2012 10:05:33 +0000 (GMT)
>>> From: olorunfemi sola <solafem7@yahoo.co.uk>
>>> Subject: Re: st: Re: working with panel data in Stata
>>>
>>> Joseph,
>>> Thank you. Your suggestion worked perfectly.
>>>
>>>
>>> ***********************************************************************
>>> SOLA OLORUNFEMI   Ph.D
>>> SENIOR LECTURER
>>>  DEAPARTMENT OF ECONOMICS
>>> AKUNGBA AKOKO
>>> ONDO STATE NIGERIA
>>> TEL NO +234 803 581 0893
>>> **********************************************************************
>>>
>>>
>>> - ----- Original Message -----
>>> From: Joseph Coveney <jcoveney@bigplanet.com>
>>> To: statalist@hsphsun2.harvard.edu
>>> Cc:
>>> Sent: Saturday, 17 March 2012, 1:26
>>> Subject: st: Re: working with panel data in Stata
>>>
>>> Try:
>>>
>>> encode id, generate(ID) label(IDs)
>>> tsset ID time, yearly
>>>
>>> Joseph Coveney
>>>
>>> P.S.  You might want to re-read Section 2.3 ("How do I send questions
>>> to
>>> Statalist?") of the list's FAQ (URL below).
>>>
>>> - -----Original Message-----
>>>
>>> Am working with panel data. The data are as follows:
>>> id    time    ep    dep    lr
>>> lagos    2000    10    11    12
>>> lagos    2001    8    12    13
>>> lagos    2002    9    9    9
>>> lagos    2003    7    10    12
>>> ondo    2000    10    8    10
>>> ondo    2001    8    12    13
>>> ondo    2002    10    11    14
>>> ondo    2003    7    10    11
>>> ife    2000    9    11    12
>>> ife    2001    8    8    11
>>> ife    2002    9    10    12
>>> ife    2003    10    9    12
>>> But when i give this command --------" tsset id time " it gave the
>>> following
>>> error:
>>>   tsset id time
>>> varlist:  id:  string variable not allowed
>>> r(109)
>>>
>>>
>>> *
>>> *   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/
>>>
>>>
>>> *
>>> *   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/
>>>
>>> ------------------------------
>>>
>>> Date: Sat, 17 Mar 2012 10:38:01 +0000
>>> From: Nick Cox <njcoxstata@gmail.com>
>>> Subject: Re: st: Parsing words on allcaps
>>>
>>> - -moss- (SSC) should help.
>>>
>>> Nick
>>>
>>> On 17 Mar 2012, at 09:53, Matthew Kerby <kerbym@mun.ca> wrote:
>>>
>>> > Dear Statalist
>>> >
>>> > Perhaps somebody can advise on how to solve this problem.
>>> >
>>> > I have a dataset which consists of a single string variable. The each
>>> > observation consists of a sentence which in turn contains a name in
>>> > all caps. I would like to create a new variable which contains only
>>> > those words which appear in all caps. eg. in example below I would
>>> > like to extract the name of the hockey player.
>>> >
>>> >
>>> > clear
>>> > input str244 string
>>> > "GRETZKY W. likes hamburgers"
>>> > "Hotdogs are preferred by LEMIEUX M."
>>> > "All things being equal, CROSBY S. will eat doughnuts"
>>> > end
>>> >
>>> > I would like to end up with a string variable (player) which looks
>>> > like the following:
>>> >
>>> > player
>>> > "GRETZKY W."
>>> > "LEMIEUX M."
>>> > "CROSBY S."
>>> >
>>> > Any help on how to parse a string variable on allcaps is greatly
>>> > appreciated.
>>> > Cheers,
>>> > M
>>> *
>>> *   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/
>>>
>>> ------------------------------
>>>
>>> Date: Sat, 17 Mar 2012 08:58:12 -0230
>>> From: Matthew Kerby <kerbym@mun.ca>
>>> Subject: Re: st: Parsing words on allcaps
>>>
>>> Thanks! Worked.
>>>
>>>
>>> On 2012-03-17, at 8:08 AM, Nick Cox wrote:
>>>
>>> > -moss- (SSC) should help.
>>> >
>>> > Nick
>>> >
>>> > On 17 Mar 2012, at 09:53, Matthew Kerby <kerbym@mun.ca> wrote:
>>> >
>>> >> Dear Statalist
>>> >>
>>> >> Perhaps somebody can advise on how to solve this problem.
>>> >>
>>> >> I have a dataset which consists of a single string variable. The
>>> each observation consists of a sentence which in turn contains a name
>>> in all caps. I would like to create a new variable which contains only
>>> those words which appear in all caps. eg. in example below I would like
>>> to extract the name of the hockey player.
>>> >>
>>> >>
>>> >> clear
>>> >> input str244 string
>>> >> "GRETZKY W. likes hamburgers"
>>> >> "Hotdogs are preferred by LEMIEUX M."
>>> >> "All things being equal, CROSBY S. will eat doughnuts"
>>> >> end
>>> >>
>>> >> I would like to end up with a string variable (player) which looks
>>> like the following:
>>> >>
>>> >> player
>>> >> "GRETZKY W."
>>> >> "LEMIEUX M."
>>> >> "CROSBY S."
>>> >>
>>> >> Any help on how to parse a string variable on allcaps is greatly
>>> appreciated.
>>> >> Cheers,
>>> >> M
>>> > *
>>> > *   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/
>>>
>>>
>>> This electronic communication is governed by the terms and conditions
>>> at
>>> http://www.mun.ca/cc/policies/electronic_communications_disclaimer_2011
>>> .php
>>>
>>> *
>>> *   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/
>>>
>>> ------------------------------
>>>
>>> Date: Sat, 17 Mar 2012 12:12:02 +0000
>>> From: "Ploy T." <ploy.tham@gmail.com>
>>> Subject: st: Two Way Fixed Effect on Unbalanced Panel Data
>>>
>>> Hi Statalist
>>>
>>>
>>> I have questions regarding two-way fixed effect on large unbalanced
>>> panel. Here is my model:
>>>
>>>
>>> ln Xijt = â0 + aij + qt + â1 lnGDPit + â2 lnGDPjt + â3 lnPOPit + â4
>>> lnPOPjt + â5  FTAijt + åijt
>>>
>>>
>>>
>>> My LHS variable is bilateral export from country i to country j. The
>>> explanatory variables are GDP and population of both countries, and a
>>> dummy variable that takes value 1 if both countries are member of an
>>> FTA. I would like to run the model for
>>> 1.) one way fixed country effect
>>> 2.) one way fixed time effect and
>>> 3.) two way fixed effect.
>>>
>>> I have unbalanced panel data for approximately 2,200 country pairs with
>>> 6 time periods. I'm a new STATA user. So, as far as I know, there is no
>>> problem on running unbalanced panel data on one-way fixed effect
>>> (correct me if I'm wromg). But there's no direct method on estimating
>>> two-way fixed effects on large unbalanced panel data. Is there any way
>>> or commands that help me get through such problem?
>>>
>>> Moreover, if I change my model such that:
>>>
>>>
>>> ln Xijt = â0 + aij +  â1 lnGDPit + â2 lnGDPjt + â3 lnPOPit + â4 lnPOPjt
>>> + â5 FTAijt - lnPit - lnPjt+ åijt
>>>
>>>
>>>
>>> That is, country-fixed effect and country-and-time fixed effects
>>> (estimating aij together with lnPit and lnPjt) are estimated together.
>>> Is there any method to get estimation of such model?
>>>
>>> Thanks very much in advance.
>>> Ploy T.
>>>
>>> *
>>> *   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/
>>>
>>> ------------------------------
>>>
>>> Date: Sat, 17 Mar 2012 09:36:59 -0400
>>> From: David Hoaglin <dchoaglin@gmail.com>
>>> Subject: Re: st: Two Way Fixed Effect on Unbalanced Panel Data
>>>
>>> Hi, Ploy.
>>>
>>> As a start, you could separate the effect aij into the "main effects"
>>> for the two countries and the interaction (e.g., replace aij with ci +
>>> cj + dij).
>>>
>>> In the second version of your model, you may want to keep the "main
>>> effect" for time (qt in the first version).
>>>
>>> You could fit both models by ordinary regression, accompanied by
>>> plotting and diagnosis, to see how the data behave.  (The number of
>>> predictors, however, may be too large for some flavors of Stata.)
>>>
>>> That regression approach does not take into account the correlation
>>> structure.  If you regard the panel as consisting of pairs of
>>> countries, a fixed-effects analysis with aij as the pair effect seems
>>> all right.  Those fixed effects for the pairs would account for
>>> structure in the data that does not change over time, and that would
>>> include the country-specific effects ci and cj that I introduced above.
>>> At the moment, I don't see how to separate aij into ci + cj + dij
>>> within a fixed-effects model, but I have not tried to search for work
>>> on panels of pairs.
>>>
>>> I hope this discussion helps.
>>>
>>> David Hoaglin
>>>
>>> 2012/3/17 Ploy T. <ploy.tham@gmail.com>:
>>> > Hi Statalist
>>> >
>>> >
>>> > I have questions regarding two-way fixed effect on large unbalanced
>>> > panel. Here is my model:
>>> >
>>> >
>>> > ln Xijt = â0 + aij + qt + â1 lnGDPit + â2 lnGDPjt + â3 lnPOPit + â4
>>> > lnPOPjt + â5  FTAijt + åijt
>>> >
>>> >
>>> >
>>> > My LHS variable is bilateral export from country i to country j. The
>>> > explanatory variables are GDP and population of both countries, and a
>>> > dummy variable that takes value 1 if both countries are member of an
>>> > FTA. I would like to run the model for
>>> > 1.) one way fixed country effect
>>> > 2.) one way fixed time effect and
>>> > 3.) two way fixed effect.
>>> >
>>> > I have unbalanced panel data for approximately 2,200 country pairs
>>> > with 6 time periods. I'm a new STATA user. So, as far as I know,
>>> there
>>> > is no problem on running unbalanced panel data on one-way fixed
>>> effect
>>> > (correct me if I'm wromg). But there's no direct method on estimating
>>> > two-way fixed effects on large unbalanced panel data. Is there any
>>> way
>>> > or commands that help me get through such problem?
>>> >
>>> > Moreover, if I change my model such that:
>>> >
>>> >
>>> > ln Xijt = â0 + aij +  â1 lnGDPit + â2 lnGDPjt + â3 lnPOPit + â4
>>> > lnPOPjt + â5 FTAijt - lnPit - lnPjt+ åijt
>>> >
>>> >
>>> >
>>> > That is, country-fixed effect and country-and-time fixed effects
>>> > (estimating aij together with lnPit and lnPjt) are estimated
>>> together.
>>> > Is there any method to get estimation of such model?
>>>
>>> *
>>> *   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/
>>>
>>> ------------------------------
>>>
>>> Date: Sat, 17 Mar 2012 14:31:52 +0000
>>> From: "Nakelse, Tebila (AfricaRice)" <T.Nakelse@cgiar.org>
>>> Subject: st: depvar name in b an V matrix
>>>
>>> Dear Statalisters,
>>> I am using regress command in Stata and I would like to know if it is
>>> possible to add the depvar name in b an V matrix as it is done in logit
>>> command.
>>> For logit  when  I do :
>>> Logit y x
>>>
>>> e(b)[1,2]
>>>         y:                       y:
>>>         x                        _cons
>>> y1  -.35129185  -2.9228238
>>>
>>> when I do
>>>  regress y x,
>>> I will have
>>>
>>> e(b)[1,2]
>>>        x                 _cons
>>> y1  -.00809294   .045464
>>>
>>> without y in e(b) column names what I want to do
>>>
>>> I know I can  pass through a regular matrix and used matrix functions
>>> such as coleq, roweq, colnames or rownames and after repost in e and V.
>>>
>>> I will appreciate any help
>>> Tebila
>>>
>>>
>>>
>>> *
>>> *   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/
>>>
>>> ------------------------------
>>>
>>> Date: Sat, 17 Mar 2012 16:23:13 +0000
>>> From: Andrea Rispoli <andrea.rspl@gmail.com>
>>> Subject: st: Variable autocorrelation
>>>
>>> Dear Statalisters,
>>>
>>> I have a panel dataset and I would like to see whether certain
>>> variables are autocorrelated to see to what extent they are
>>> slow-moving versus fast-moving. What kind of test/ command can I
>>> run?This is perhaps a question with a well known answer but I could
>>> not find an answer after searching for a while so if someone my give
>>> me a suggestion I would really appreciate it!
>>> *
>>> *   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/
>>>
>>> ------------------------------
>>>
>>> Date: Sat, 17 Mar 2012 18:02:51 -0400
>>> From: Amanda Fu <mandy.fu1@gmail.com>
>>> Subject: st:AR(1) error notice: sample may not include multiple panels
>>>
>>> Dear Statalists,
>>>
>>> Sorry for bothering you with  a question about a simple AR(1)
>>> estimation. I searched related discussion in the archives , but still
>>> cannot figure out why the error message comes out. Any suggestions
>>>
>>> I want to estimate a univarate AR(1) model (without any controls). The
>>> data set is a panel data,  including 1000 observations for 5 years
>>> (200 variables. The one I am interested is enroll taking values 0,1) .
>>> . tsset id years
>>> . arima enrollment,ar(1)
>>> "sample may not include multiple panels"
>>>
>>> I must have missed something here. Can someone give me some hints?
>>> In addition, instead of using ARIMA, may I know if there is any easy
>>> alternative to do the AR(1) ?
>>>
>>> Thank you very much!
>>>
>>> Amanda
>>> *
>>> *   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/
>>>
>>> ------------------------------
>>>
>>> Date: Sun, 18 Mar 2012 01:06:57 +0000
>>> Subject: Re: st:AR(1) error notice: sample may not include multiple
>>> panels
>>>
>>> Amanda Fu wrote:
>>>
>>> > Sorry for bothering you with  a question about a simple AR(1)
>>> > estimation. I searched related discussion in the archives , but still
>>> > cannot figure out why the error message comes out. Any suggestions
>>> >
>>> > I want to estimate a univarate AR(1) model (without any controls).
>>> The
>>> > data set is a panel data,  including 1000 observations for 5 years
>>> > (200 variables. The one I am interested is enroll taking values 0,1)
>>> .
>>> > . tsset id years
>>> > . arima enrollment,ar(1)
>>> > "sample may not include multiple panels"
>>> >
>>> > I must have missed something here. Can someone give me some hints?
>>> > In addition, instead of using ARIMA, may I know if there is any easy
>>> > alternative to do the AR(1) ?
>>>
>>> ARIMA models are designed for _single_ time series, not panel data.
>>> Since you're looking to specify AR(1) effects, that also rules out
>>> fixed- and random-effect models under -xtreg-, since parameter
>>> estimates are almost always biased and inconsistent in the presence of
>>> AR(1). Your only real alternative is investigate instrumental variable
>>> models, particularly IV-2SLS (-h xtivreg-), assuming you can find
>>> suitable variables to 'instrument' with the lagged variable. That's
>>> where the fun starts, although I use the word 'fun' advisedly.
>>>
>>> - --
>>> Clive Nicholas
>>>
>>> [Please DO NOT mail me personally here, but at
>>> <clivenicholas@hotmail.com>. Please respond to contributions I make in
>>> a list thread here. Thanks!]
>>>
>>> "My colleagues in the social sciences talk a great deal about
>>> methodology. I prefer to call it style." -- Freeman J. Dyson
>>>
>>> *
>>> *   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/
>>>
>>> ------------------------------
>>>
>>> Date: Sat, 17 Mar 2012 21:13:21 -0400
>>> From: Richard Herron <richard.c.herron@gmail.com>
>>> Subject: Re: st:AR(1) error notice: sample may not include multiple
>>> panels
>>>
>>> This question came up yesterday, too. Here was my response:
>>>
>>> Panels require strict exogeneity for consistency, so you can't have
>>> lagged dependent variables on the right hand side (i.e., dynamic
>>> panels aren't allowed).
>>>
>>> If you think about the within estimator for panel data, then you have
>>> y_{i, t-1} - mean(y_i) on the right hand side, which is not orthogonal
>>> to e_{i,t} because mean(y_i) contains y_{i,t} from the left hand side.
>>>
>>> There are some solutions to this in -xtabond-, which does the Arellano
>>> and Bond estimator. Arellano and Bond correct the endogeneity by
>>> estimating the dynamic panel in first differences and using lags as
>>> instruments and GMM to take care of over-identification.
>>>
>>> Chapter 9 in Cameron and Trivedi's "Microeconometrics Using Stata"
>>> provides a handful of techniques for estimating dynamic panel models.
>>> Chapters 21 and 22 in their Microeconometrics textbook provides more
>>> theory. HTH.
>>>
>>> On Sat, Mar 17, 2012 at 18:02, Amanda Fu <mandy.fu1@gmail.com> wrote:
>>> > Dear Statalists,
>>> >
>>> > Sorry for bothering you with  a question about a simple AR(1)
>>> > estimation. I searched related discussion in the archives , but still
>>> > cannot figure out why the error message comes out. Any suggestions
>>> >
>>> > I want to estimate a univarate AR(1) model (without any controls).
>>> The
>>> > data set is a panel data,  including 1000 observations for 5 years
>>> > (200 variables. The one I am interested is enroll taking values 0,1)
>>> .
>>> > . tsset id years
>>> > . arima enrollment,ar(1)
>>> > "sample may not include multiple panels"
>>> >
>>> > I must have missed something here. Can someone give me some hints?
>>> > In addition, instead of using ARIMA, may I know if there is any easy
>>> > alternative to do the AR(1) ?
>>> >
>>> > Thank you very much!
>>> >
>>> > Amanda
>>> > *
>>> > *   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/
>>>
>>> *
>>> *   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/
>>>
>>> ------------------------------
>>>
>>> Date: Sat, 17 Mar 2012 21:34:48 -0400
>>> From: David Hoaglin <dchoaglin@gmail.com>
>>> Subject: Re: st: ztnb model: why are some results omitted?
>>>
>>> Dear Ekaterina,
>>>
>>> How are you interpreting the constant term in the model?  Because the
>>> four income categories are exhaustive, they need only three parameters
>>> in addition to the constant term.  It seems likely that ztnb is
>>> treating the highest quartile of income as the reference category.
>>>
>>> I suggest that you merge the four income variables into a single
>>> categorical variable (which you might name inccat) and then use
>>> inccat##conf_inc as the list of independent variables.
>>>
>>> Regards,
>>>
>>> David Hoaglin
>>>
>>> On Fri, Mar 16, 2012 at 6:14 PM, Ekaterina Hertog
>>> <ekaterina.hertog@sociology.ox.ac.uk> wrote:
>>> > Dear all,
>>> > I use Stata 12 and I am trying to analyse how individual relative
>>> income and
>>> > whether the level of income was supported by official documentation
>>> or not
>>> > affects the number of page views an person receives on a dating site.
>>> I am
>>> > particularly interested in the interactions between the relative
>>> income
>>> > variables and the variable for confirmed income.
>>> > I have no information about people who receive no page views so I use
>>> a
>>> > zero-truncated negative binominal model.
>>> > The dependent variable is a count of the number of page views.
>>> > inclow25 = income is in the lowest quartile of incomes in the general
>>> > population
>>> > inc50per = income in the 2nd lowest quartile
>>> > inc75per = income in the 2nd highest quartile
>>> > inchigh25 = income in the highest quartile of earners in the
>>> population
>>> > conf_inc = is a binary variable: 1 meaning that the income level is
>>> > confirmed with appropriate documentation
>>> > I run the model separately for men and the code I use looks as
>>> follows:
>>> > ztnb totpagev inclow25##conf_inc inc50per##conf_inc
>>> inc75per##conf_inc
>>> > inchigh25##conf_inc  if gender==1
>>> >
>>>
>>> *
>>> *   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/
>>>
>>> ------------------------------
>>>
>>> Date: Sat, 17 Mar 2012 23:12:14 -0400
>>> From: Amanda Fu <mandy.fu1@gmail.com>
>>> Subject: Re: st:AR(1) error notice: sample may not include multiple
>>> panels
>>>
>>> Hello Mr. Nicholas and Mr. Herron,
>>>
>>> Thank you very much for your suggestions.
>>>
>>> After seeing the above suggestions, I checked chapter 9 in Cameron and
>>> Trivedi again and saw there was some useful discussion. I am sorry
>>> that I missed this this morning. I focused on searching stuff related
>>> to "time series" in that book.
>>>
>>> In addition to Arellano-Bond estimator, would it be fine if I use
>>> Cochrane–Orcutt estimation (-prais, cort) to estimate AR(1)?
>>>
>>> Thank you!
>>> Amanda
>>>
>>> On Sat, Mar 17, 2012 at 9:13 PM, Richard Herron
>>> <richard.c.herron@gmail.com> wrote:
>>> > This question came up yesterday, too. Here was my response:
>>> >
>>> > Panels require strict exogeneity for consistency, so you can't have
>>> > lagged dependent variables on the right hand side (i.e., dynamic
>>> > panels aren't allowed).
>>> >
>>> > If you think about the within estimator for panel data, then you have
>>> > y_{i, t-1} - mean(y_i) on the right hand side, which is not
>>> orthogonal
>>> > to e_{i,t} because mean(y_i) contains y_{i,t} from the left hand
>>> side.
>>> >
>>> > There are some solutions to this in -xtabond-, which does the
>>> Arellano
>>> > and Bond estimator. Arellano and Bond correct the endogeneity by
>>> > estimating the dynamic panel in first differences and using lags as
>>> > instruments and GMM to take care of over-identification.
>>> >
>>> > Chapter 9 in Cameron and Trivedi's "Microeconometrics Using Stata"
>>> > provides a handful of techniques for estimating dynamic panel models.
>>> > Chapters 21 and 22 in their Microeconometrics textbook provides more
>>> > theory. HTH.
>>> >
>>> > On Sat, Mar 17, 2012 at 18:02, Amanda Fu <mandy.fu1@gmail.com> wrote:
>>> >> Dear Statalists,
>>> >>
>>> >> Sorry for bothering you with  a question about a simple AR(1)
>>> >> estimation. I searched related discussion in the archives , but
>>> still
>>> >> cannot figure out why the error message comes out. Any suggestions
>>> >>
>>> >> I want to estimate a univarate AR(1) model (without any controls).
>>> The
>>> >> data set is a panel data,  including 1000 observations for 5 years
>>> >> (200 variables. The one I am interested is enroll taking values 0,1)
>>> .
>>> >> . tsset id years
>>> >> . arima enrollment,ar(1)
>>> >> "sample may not include multiple panels"
>>> >>
>>> >> I must have missed something here. Can someone give me some hints?
>>> >> In addition, instead of using ARIMA, may I know if there is any easy
>>> >> alternative to do the AR(1) ?
>>> >>
>>> >> Thank you very much!
>>> >>
>>> >> Amanda
>>> >> *
>>> >> *   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/
>>> >
>>> > *
>>> > *   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/
>>>
>>> *
>>> *   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/
>>>
>>> ------------------------------
>>>
>>> End of statalist-digest V4 #4460
>>> ********************************
>>>
>>> *
>>> *   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/
>>
>> *
>> *   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/
>
> *
> *   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/

*
*   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/
```