Re: st: merge creates duplicates in master data

[Will Hauser <whauseriii@gmail.com>]

Jose,
I appreciate your help but could you be a bit more clear about the 
joinby command?  Joinby results in many more duplicates than I obtain 
with the merge command.  The unmatched option must be specified as none 
(the default), master, using, or both.  I tried 'none' and 'master', the 
two plausible specifications but both result in several hundred duplicates.

My primary concern with the merge command is not so much the resulting 
duplicates - they are few and I can identify and remove them with the 
'duplicates' command.  Rather, it is that I do not fully understand the 
command itself or there is some flawed logic in my multi-step approach.  
Additional documentation on the merge command or any other insights as 
to how I might best match these lists would be most helpful.

One tidbit I did not add in the first post is that in addition to the 
update option I am also using the nokeep option.  The master list is the 
definitive list if an observation is not on that list then that 
observation will not be needed.

I may try to track down a copy of stata11 to try the new merge command.  
Thanks for that suggestion.

Will


Jose Ricardo Nogueira wrote:
> Stata 11 has a new merge command that will get rid of this problem,
> but in Stata 10 you can do this with joinby command and unmatched
> option.
>
> Sincerely,
>
> Jose
>
>   
> > Hello all,
> >
> > I am experiencing unexpected behavior in Stata 10 when using the merge
> > command.
> >     
> *
> *   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/