I can't help with the specifics of -xml_tab-, which I have not used
or examined very closely.
Rather, I can comment on the use of -version- here.
Generally, -version- is not a time machine, or even a version machine.
If you want Stata 8 behaviour, for example, the only surefire way to get
that is to use Stata 8.
Also, -version- does much less than some people hope it will do. By and
large, it sets things back only in respect of things documented under
-version-. I think that's not the whole truth, but it's nearer the whole
truth than the idea that -version- is a time machine.
As -xml_tab- is already written for Stata 8, I wouldn't expect using
-version 8- in calling it would make any difference, given that there is
no use within -xml_tab- of -_caller()-.
Nick
[email protected]
Daniel Klein
Thanks for your help. I have the same xml_tab version. However, when I
run your program it still generates an error:
/ln_sig invalid name
r(198);
I noticed that the ado was written for version 8. My version is 10. So I
tried
vers 8: xml_tab, replace save(c:\test)
, which most times generated the same error message. Sometimes it worked
but I could not infer the reason for that. The version command works
strange...
-Daniel
Zurab Sajaia schrieb:
> Hi Daniel,
>
> I have -xml_tab- version 3.35 17Oct2007 available from SSC
> and just tried following:
>
> sysuse auto, clear
> stset length
> streg price rep78 trunk turn, distribution(gamma)
>
> xml_tab, replace save(c:\test)
>
> it worked.
>
> Could you give me the full command line you use? if you have the same
> version but still get an error, there might be a case that it happens
> only for specific combination of options.
>
> Best,
> Zurab
>
>
> ----- Original Message ----- From: "Daniel Klein"
> <[email protected]>
> To: <[email protected]>
> Sent: Tuesday, January 29, 2008 4:43 AM
> Subject: st: "xml_tab" and "streg"
>
>
>> Hi,
>>
>> I really like xml_tab which is a usefull tool to export tables of
>> estimates directly to Excel. However when I estimate a survival model
>> with streg and weibull distribution I get the following error
message.
>>
>> /ln_p invalid name
>> r(198)
>>
>> This is not the case when I assume an exponential distribution since
in
>> that case the parameter "p" (or ln_p) is not estimated. I guess
xml_tab
>> has a problem with the "/". estout works fine but I would more like
to
>> further use xml_tab.
>>
>> Does anyone know a workaround? Thanks in advance.
*
* 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/
