Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: Panel data and poisson-regression


From   Austin Nichols <austinnichols@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: Panel data and poisson-regression
Date   Thu, 31 Dec 2009 13:44:30 -0500

Clarify by worked example, or carefully chosen words, not confused
pseudocode (note that -xtline- applied to the time variable produces a
bunch of y=x function graphs).

webuse psidextract, clear
collapse (sum) _freq=union, by(ed t)
ren ed state
tsset state t
qui xi: poisson _freq t i.state
predict n
xtline n _freq if state>12

On Thu, Dec 31, 2009 at 1:12 PM, moleps <moleps2@gmail.com> wrote:
> What I want is to overlay the poisson predicted values (that I´m unsure how to get), ie :
>
> tsset state _freq
> xtline _freq
> xtpoisson ...something...
> predict pr
>
> xtline ***_freq & pr (/overlaid on each other**
>
> Happy new year,
>
> M
>
> On 31. des. 2009, at 19.02, Martin Weiss wrote:
>
>> <>
>>
>> What exactly do you want to overlay? The predicted number from the -poisson- regressions? Please clarify.
>>
>>
>> Happy new year
>> Martin
>> -------- Original-Nachricht --------
>>> Datum: Thu, 31 Dec 2009 18:14:59 +0100
>>> Von: moleps <moleps2@gmail.com>
>>> An: statalist@hsphsun2.harvard.edu
>>> Betreff: st: Panel data and poisson-regression
>>
>>> Dear listers,
>>> I´m looking for an effect over time between different states on the
>>> incidence of a rare cancertype. I´ve contracted my data so that I´m only left
>>> with _freq, state and year.
>>>
>>> .tsset state _freq
>>> .xtline _freq
>>>
>>> and from here on I´m stuck. What I´d like to do is to fit a poisson
>>> regression model for each of the states and have it overlaid on the -xtline
>>> _freq-...
>>>
>>> Any ideas on how to do this?
>>>
>>> regards,
>>> 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/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index