I don't see that anyone mentioned that it is a bad
idea to use a scalar for this purpose, because
scalars and variables share the same name space.
Here is a dopey example.
. sysuse auto
(1978 Automobile Data)
. scalar mpg = 10
. forval i = 1/`=mpg' {
2. di `i'
3. }
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
And another
. scalar m = 10
. forval i = 1/`=m' {
2. di `i'
3. }
m ambiguous abbreviation
invalid syntax
r(198);
Asked to look for -mpg-, Stata uses the value -mpg[1]-.
Asked to look for -m-, Stata can't decide whether you want
-mpg[1]- or -make[1]-.
This in practice might not bite very often, but when it does, it
could prove very puzzling and waste a lot of your time.
The problem is all documented, but quite simply, don't do this.
Use locals instead!
Nick
[email protected]
[email protected]
There might be some typos in the previous code I said it worked.
I would like to assure you that the = is required and it works:
. forval i=1/`=ng' {
2. mat p=inv((r1+vu)/ns[`i'])*r1
3. mat q=r2*(idd-p)
4. mat xtilde`i' = xmean*p +mux3*(idd-p)+dz*q
5. }
*
* 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/
