Radu Ban
>
> I have the following loop (this is only a part of a do file)
> .....
>
> for any prad_coeff prad_error prad_tval constant_coeff
> constant_error
> constant_tval nrobs nonzeroobs regnum: gen X = 999;
>
> local i = 1;
>
> foreach var of varlist `outcomes' {;
>
> reg `var' prad;
>
> mat C1`i' = e(b);
>
> mat E1`i' = e(V);
>
> replace prad_coeff[`i'] = C1`i'[1,1];
< snip >
>
> replace count if `var' ~= 0 & `var' ~= .;
>
< snip>
>
> weights not allowed
>
> I'm not sure what this means, but I suspect that I cannot
> replace variables
> observation by observation. Is there a way to work around this?
>
As Tao Jiang wrote
> if prad_coeff is variable, use:
>
> replace prad_coeff = C1`i'[1,1] in `i'
>
> dp similar changes for others.
The error message arises because
Stata guesses that you are trying to specify weights
and weights, specified in [ ], are not allowed on the LHS of
a -replace- statement. It is also true that subscripts
specified in [] are not allowed on the LHS of a -replace-
statement, but then they are not allowed in any Stata statement
that I can think of.
Your -replace count- will fail, in the
absence of an expression.
Nick
[email protected]
*
* 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/
