Hi,
I have a large set of short time-series (one for each patient) of about 2000 patients. The patient is identified by a record number rc and the time series data point by pnt. I have used
sort rc pnt
by rc : regress dt lv
to obtain a lin. regression for each patient (as an example - the real thing will use non-linear regression).
I am interested in the results of all these little regressions (coefficients, P values and Rsquared). My question is - can I extract these from the regression results (which go to the screen) in order to put them in a new variable. The aim is to summarise each patient's time series using the results of their regression model.
Like this:
Patient Rsquare const P coef P
Is this possible? Or do I have to write the regression method explicitly (ie not using regress)?
Thanks to all.
Keith Farnsworth (Epidemiologist - Dundee University, Scotland).
*
* 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/
