The last line would be better as
by cy: replace kstck = 0.93 * L.kstck + L.gfc if _n > 1
Nick
[email protected]
-----Original Message-----
From: [email protected]
[mailto:[email protected]] On Behalf Of Nick Cox
Sent: 08 September 2009 12:42
To: [email protected]
Subject: st: RE: about 'forvalues' loop
0. A problem that will bite
============================
The syntax `i-1' won't work. It would need to be `=`i' - 1'
That is legal, but in your case it looks as if you have various other
problems that affect your code earlier.
1. A problem that won't bite
============================
In
bysort cy: gen kstck = 0
the -bysort- does no harm, but it does nothing. The result is 0,
regardless of -cy-.
2. A general problem
====================
Reshaping to wide will on balance make your life _much_ more difficult
here.
It looks as if a better approach would be something like
gen kstck = 0
tsset cy yr
by cy: replace kstck = 0.93 * L.kstck + L.gfc
Nick
[email protected]
Marwan Elkhoury
sorry again to bother you with some very stupid question but I'm
getting stuck again on some simple matter that I believe is in the
writing of scalar.
the story is that I have a panel data and I'ld like to generate a new
variable called kstck (by country (cy)).
so I run this loop:
bysort cy: gen kstck = 0
reshape wide gdp gfc kstck, i(cy) j(yr)
* Calculate capital stock from 1990 to 2000 using the following
formula (where gfc* is the investment in year *)
bysort cy: replace kstck1990 = gfc1990
forvalues i = 1991/2000 {
bysort cy: replace kstck`i' = .93*kstck`i-1' + gfc`i'
}
However, it is not giving me the required result of the recursive
summation. why ?
*
* 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/
