Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: RE: Creating a dummy variable that 'marks out' useless cases


From   "Nick Cox" <[email protected]>
To   <[email protected]>
Subject   st: RE: Creating a dummy variable that 'marks out' useless cases
Date   Mon, 13 Dec 2004 15:01:25 -0000

In addition to Scott Merryman's post, I'll 
observe that indicator questions are often 
easier flipped the other way up, just like
probability questions. 

Positively, Clive is happy with 

conch < . & edconch < . & marker != 0 

and so 

gen mark = conch < . & edconch < . & marker != 0 

should be close to the mark. 

Nick 
[email protected] 

Clive Nicholas

> I'm trying to -egen-erate a dummy variable which 'marks off' 
> cases in my
> dataset which are both (1) missing on two key variables and 
> (2) have been
> already been coded 0 as duplicate cases by an -egen, 
> tag()-ged variable
> called -marker-. I want the generated variable (-mark-) to 
> look like this:
> 
>       +-----------------------------------------------+
>       | pano       conch     edconch   marker    mark |
>       |-----------------------------------------------|
> 1600. |  436     2.08482    1.042244        1       1 |
> 1601. |  436   -2.551552   -23.68557        0       0 |
> 1602. |  436    1.171085   -18.44246        0       0 |
> 1603. |  436   -2.551552   -28.24039        0       0 |
> 1604. |  436    1.171085   -23.11701        0       0 |
>       |-----------------------------------------------|
> 1605. |  436    9.708897    -6.17593        0       0 |
> 1606. |  436    9.708897   -9.517298        0       0 |
> 1607. |  437   -2.786781           .        1       0 |
> 1608. |  437   -.6094589           .        1       0 |
> 1609. |  437   -12.96428           .        1       0 |
>       |-----------------------------------------------|
> 1610. |  437    1.095558           .        1       0 |
> 1611. |  438    7.136993    21.97673        1       1 |
> 1612. |  438    .5475883    5.041264        1       1 |
> 1613. |  438    2.488335    .6113586        1       1 |
> 1614. |  438   -6.498318    -11.1867        1       1 |
>       |-----------------------------------------------|
> 1615. |  438    .7604446    21.50825        1       1 |
> 1616. |  439    6.331062    1.482658        1       1 |
> 1617. |  439   -1.661823   -11.79961        1       1 |
> 1618. |  439    2.188976   -3.028099        1       1 |
> 1619. |  439   -6.502392   -17.35701        1       1 |
>       |-----------------------------------------------|
> 1620. |  439    3.512476    8.189182        1       1 |
>       +-----------------------------------------------+
> 
> As you can see, -mark- is 0 when _either_ -conch- or 
> -edconch- is missing
> _or_ when -marker- is 0 (or, indeed, both). I thought this 
> would be a dead
> simple task (perhaps it really is), but all the commands I've 
> issued so
> far aren't executing what I want. For example:
> 
> (1) Trying to do it in two:
> 
>     gen mark=0 if conch!=. & edconch!=. & marker==1
>     recode mark .=1
> 
> (2) Trying to do it step-by-step:
> 
>     gen mark=0 if marker==0
>     replace mark=0 & conch==.
>     replace mark=0 if edconch==.
>     replace mark=1 if marker==1
>     replace mark=1 & conch!=.
>     replace mark=1 if edconch!=.
> 
> (3) Using -egen- functions (admittedly, partly out of desperation):
> 
>     egen mark=eqany(conch edconch) if marker==1 & conch!=. //
>     & edconch!=., values(-100/100)
> 

*
*   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/



© Copyright 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index