Dear Kit and Dick,
Thank you for your comments. In fact, I was working on my own hacked
version of xtpattern, since I also preferred to transform the string 1.1 to
101 and then to binary. However, my dataset is of the form:
1111...............
.1111.............
..1111............
...1111.......... etc.
so the binary trick was not so useful (I may want to discard a
1..................
but keep a ..................1111, and the first number is still greater).
As Kit points out, the shift-bit part of it was not very good with stata.
I will try the hacked version of xtpattern - fortunately I am not missing a
whole
year at any point. But if this is the case, maybe a fillin would help? or may be
(not a very elegant solution), just creating one false observation for the
missing year
could help...
Thank you very much again
The "hack":
copy http://fmwww.bc.edu/ec-p/baum/xtpattern2.ado xtpattern2.ado
ssc install spell
***************************
Guillermo Cruces
*
* 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/
