RE: st: repeated measures

[Nick Cox <n.j.cox@durham.ac.uk>]

Perhaps. But David Bell's suggestion of -collapse- I think trumps all other suggestions. 

Nick 
n.j.cox@durham.ac.uk 

Michael McCulloch

Thanks Nick, you're correct I did misread that part.
Nevertheless, I suspect that some of the complexity you allude to your  
your prior reply to Michael E. is that he wishes to selectively carry  
forward certain field values. For example, something like:
	replace the missing value from (first visit at which it's available )  
if that value is present in that patient's first visit
I would know how to define the first visit a patient had, but not how  
to find the first visit at which a lab value appeared.

On Sep 7, 2010, at 1:12 PM, Nick Cox wrote:

> I think Michael M. is misreading this. It's a long to long mapping,  
> so -reshape- does not spring to mind here. See my earlier suggestions.

Michael McCulloch

> This is possible with the -reshape- command. What you are starting
> with is the "long" data format, and you can use that command to change
> to "wide." However, there appear to be a few problems with your
> description:
> 1. Did you mean patient ID=4, not =3, in the last line of the data
> format you're seeking?
> 2. In the desired data format you list the "glu" data point for
> subject ID=3 as being from a different date than the original 6/1/09.
>
> On Sep 7, 2010, at 12:54 PM, Michael Eisenberg wrote:
>
>> I have a problem I hope you can help with that involves cleaning a
>> dataset for analysis.
>>

*
*   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/