Thanks Nick
Works like a charm (can it be generalized for similar problems or is it not
worth the extra bit)
Ronnie
Ronnie wrote
> (Stata 8, win 2K) I have a correlation matrix in excel (I
> dont have the data
> from which it was obtained) and I would like to perform
> factor analysis on
> this matrix. This is possible if by using corr2data, I generate the
> underlying data and then I can perform the factor analysis.
>
> Here is my correlation matrix
> Corr Matrix
> 1
> 0.439 1
> 0.41 0.351 1
> 0.288 0.354 0.164 1
> 0.329 0.32 0.19 0.595 1
> 0.248 0.329 0.181 0.47 0.464 1
>
> If I try to read this in stata (using insheet), I get 6 new
> variables with
> missing cases
>
> +---------------------------------------+
> | v1 v2 v3 v4 v5 v6 |
> |---------------------------------------|
> | 1 . . . . . |
> | .439 1 . . . . |
> | .41 .351 1 . . . |
> | .288 .354 .164 1 . . |
> | .329 .32 .19 .595 1 . |
> |---------------------------------------|
> | .248 .329 .181 .47 .464 1 |
> +---------------------------------------+
> I then convert this into a matrix using mkmat and I get
>
> a[6,6]
> 1 . . . . .
> .43900001 1 . . . .
> .41 .35100001 1 . . .
> .28799999 .354 .164 1 . .
> .329 .31999999 .19 .59500003 1 .
> .248 .329 .18099999 .47 .46399999 1
>
> Question. How can I tell stata that this is a symetric
> matrix and fill in
> the missing ijs with the corresponding ji values? (or, how
> can I do this in
> excel)
forval i = 1/6 {
forval j = 1 / 6 {
if a[`i',`j'] == . matrix a[`i',`j'] = a[`j',`i']
}
}
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/
*
* 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/
