Re: st: RE: RE: RE: ci, return lists, and statsby commands

[Fred Wolfe <fwolfe@arthritis-research.org>]

Roger, is there a simple way to obtain exponentiated estimates/95% CI using 
parmsby and xtgee?

Thanks,

Fred



> > Here's one way. Another way would be to
> > do -statsby- separately for each variable
> > and -merge- the results. Another way
> > would be to -stack- your responses into
> > one variable and then apply -statsby-.
> > There are, no doubt, yet other ways.
> One possible other way might be to use -parmby- (part of tthe -parmest- 
> package) instead of -statsby-, and to use the -regress- command to 
> generate the means for each variable, and then to use -dsconcat- to 
> concatenate the output data sets for each variable. For instance, John 
> might write a program
>
> local i1=0
> foreach X of var price mpg weight length {
>   local i1=`i1'+1
>   tempfile tf`i1'
>   parmby "regress `X'", saving(`tf`i1',replace) ylabel
> }
> dsconcat `tf1' `tf2' `tf3' `tf4'
> gene foldrange=max95/min95
>
> This will create a data set in the memory with 1 observation per variable 
> and data on parameter estimates, confidence limits, P-values and fold 
> ranges, overwriting the existing data. Optionally, the user can also save 
> other estimation results, such as sample numbers.
>
> To find out more about the -parmest- and -dsconcat- packages, type -ssc 
> desc parmest- or -ssc desc dsconcat- inside Stata. More information about 
> these and other packages for post-processing results can be found in 
> Newson (2003), which can be downloaded as a pre-publication draft from my 
> website at
> http://www.kcl-phs.org.uk/rogernewson/
> either using a browser or using the Stata -net- command.
>
> I hope this helps.
>
> Roger




Fred Wolfe
National Data Bank for Rheumatic Diseases
Wichita, Kansas
Tel (316) 263-2125     Fax (316) 263-0761
fwolfe@arthritis-research.org


*
*   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/