Thanks Nick, ur suggestion solves my problem.
-----Original Message-----
From: [email protected]
[mailto:[email protected]]On Behalf Of Nick Cox
Sent: 4. februar 2003 14:41
To: [email protected]
Subject: st: RE: replacing missing cases
Ronnie Babigumira
> I am running regressions on sub-samples (diffrent years) of
> a data set. I
> would like to add the crimerate (lcrmrte) of 1986 to a
> regression for 1987
> so I generate lcrmte86 which is the rate for 1986 however
> as u can see, its
> missing if the year is not 1986 (which makes sense). I
> would now like to
> fill in the missing values with the county value for 1986
> (such that lcrmte
> = -3.359507 when county == 1 and so forth).
>
> How do I replace the missing cases in lcrmte86 with the
> county value given.
>
> lcrmrte lcrmte86 county year
> -3.221757 . 1 81
> -3.261134 . 1 82
> -3.496449 . 1 83
> -3.36027 . 1 84
> -3.308445 . 1 85
> -3.359507 -3.359507 1 86
> -3.335309 . 1 87
> -4.110956 . 3 81
> -3.959896 . 3 82
> -4.189807 . 3 83
> -4.29313 . 3 84
> -4.41997 . 3 85
> -4.342945 -4.342945 3 86
> -4.182966 . 3 87
> -4.673753 . 5 81
> -4.396296 . 5 82
> -4.35781 . 5 83
> -4.280922 . 5 84
> -4.384005 . 5 85
> -4.251806 -4.251806 5 86
> -4.345864 . 5 87
> -3.820543 . 7 81
> -4.052395 . 7 82
> -3.720948 . 7 83
> -3.801224 . 7 84
> -3.705627 . 7 85
> -3.724378 -3.724378 7 86
> -3.621101 . 7 87
>
One way to do it:
. bysort county (lcrmte86) : replace lcrmte86 = lcrmte86[1] if
missing(lcrmte86)
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/
