Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down on April 23, and its replacement, statalist.org is already up and running.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: how to generate parent variables matched to their children in household level data set?


From   Nick Cox <njcoxstata@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: how to generate parent variables matched to their children in household level data set?
Date   Sat, 23 Feb 2013 09:33:13 +0000

I am at a loss to understand what you are asking. My previous posts
showed that with your sample data the code I used does work. It
remains a mystery why you first reported otherwise, and also why you
imply that the problem you stated is still unsolved. I just did that
for you. It seems that you have not studied my code and its results.

The absence of a single clear indicator variable is immaterial here.
You want to copy data from mothers' and fathers' observations to
children's; for that being able to link mother and father identifiers
to children is necessary and sufficient, and done separately.

My mention of -merge- just hints at a different method, but I have
given a method that works. I was not stating or implying that you need
to -merge-; that's merely a good alternative.

If you want to know why my method works you need to study not only
discussion of loops as in

SJ-2-2  pr0005  . . . . . .  Speaking Stata:  How to face lists with fortitude
        . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  N. J. Cox
        Q2/02   SJ 2(2):202--222                                 (no commands)
        demonstrates the usefulness of for, foreach, forvalues, and
        local macros for interactive (non programming) tasks

but also the use of -by:- as in

SJ-2-1  pr0004  . . . . . . . . . . Speaking Stata:  How to move step by: step
        . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  N. J. Cox
        Q1/02   SJ 2(1):86--102                                  (no commands)
        explains the use of the by varlist : construct to tackle
        a variety of problems with group structure, ranging from
        simple calculations for each of several groups to more

My code requires the fact that under the aegis of -by:-  subscripts
(42 in -foo[42]- is a subscript) are numbered within groups, so the
subscript [1] refers to the first observation in each group.

As said, I don't see that you need any further code, so I have not
studied your code beyond noticing that -forevar- is not a Stata
command.

Nick

On Sat, Feb 23, 2013 at 8:36 AM, Haena Lee <hannahlee419@gmail.com> wrote:
> Nick,
>
> I would love to merge father's and mother's data with children. That
> was my first choice.
> As you may have noticed, however, my data doesn't have one clear
> indicator variable of who is mother/father/child/grandparent. Although
> there are ID_F and ID_M,  what makes me confused is, ID_F and ID_M are
> on the same row of  children. I see "fid and mid" from your previous
> answer is also located on children's row. So how do I tell stata to
> generate a new indicator of "mothers" and to treat it as a property of
> mothers, not children? So that eventually I would extract moms from
> this raw data (e.g., keep ID BMI_M EMP_M if mom==1) and merge (1:many)
> it based on key variable (ID_fam) with children's data?
>
> Assuming looping would do this work,
>
> gen mom=.
> unab Y: ID
> unab Z: ID_M
> forevar x of newlist mom
>         replace `x' ==1 if Y==Z
>  }
>
> Please note that I am not familiar with the concept of looping. Just
> taught myself today for a little bit so I am not sure if those
> commands above would make sense. If not, let me know. I'd happy to
> explain it again.
>
> Haena
>
> On Fri, Feb 22, 2013 at 7:54 PM, Nick Cox <njcoxstata@gmail.com> wrote:
>> Note that I wrote that FAQ some years ago. Now I think why didn't I
>> approach that as a -merge- problem?  Create a dataset with fathers'
>> data, one with mothers' data, and -merge- using those. There is still
>> some fiddling around. This all goes with the simple idea that we have
>> favourite tools.
>>
>> Nick
>>
>> On Sat, Feb 23, 2013 at 1:50 AM, Nick Cox <njcoxstata@gmail.com> wrote:
>>> That's an allusion is to my FAQ
>>>
>>> FAQ     . . Creating variables recording prop. of the other members of a group
>>>         . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  N. J. Cox
>>>         4/05    How do I create variables summarizing for each
>>>                 individual properties of the other members of a
>>>                 group?
>>>
>>> http://www.stata.com/support/faqs/data-management/creating-variables-recording-properties/
>>>
>>> I don't know why you report problems. The code suggested there works
>>> as intended. Here it is again run on your example data:
>>>
>>> . by ID_fam (ID), sort: gen pid = _n
>>>
>>> . gen byte fid = .
>>> (7 missing values generated)
>>>
>>> . gen byte mid = .
>>> (7 missing values generated)
>>>
>>> . summarize pid, meanonly
>>>
>>> . forval i = 1 / `r(max)' {
>>>   2.                 by ID_fam: replace fid = `i' if ID_F == ID[`i'] &
>>> !missing(ID_F)
>>>   3.                 by ID_fam: replace mid = `i' if ID_M == ID[`i'] &
>>> !missing(ID_M)
>>>   4. }
>>> (3 real changes made)
>>> (0 real changes made)
>>> (0 real changes made)
>>> (3 real changes made)
>>> (0 real changes made)
>>> (0 real changes made)
>>> (0 real changes made)
>>> (0 real changes made)
>>>
>>> . l
>>>
>>>      +----------------------------------------------------------------------------------+
>>>      |       ID_F         ID_M      BMI           ID     ID_fam   Emp
>>>  pid   fid   mid |
>>>      |----------------------------------------------------------------------------------|
>>>   1. |                           26.501   A901963701   A9019637     1
>>>    1     .     . |
>>>   2. |                           20.483   A901963702   A9019637     1
>>>    2     .     . |
>>>   3. | A901963701   A901963702   20.924   A901963703   A9019637     .
>>>    3     1     2 |
>>>   4. |                           27.209   A901963801   A9019638     1
>>>    1     .     . |
>>>   5. |                           31.733   A901963802   A9019638     .
>>>    2     .     . |
>>>      |----------------------------------------------------------------------------------|
>>>   6. | A901963801   A901963802   18.018   A901963803   A9019638     .
>>>    3     1     2 |
>>>   7. | A901963801   A901963802   19.054   A901963804   A9019638     .
>>>    4     1     2 |
>>>      +----------------------------------------------------------------------------------+
>>>
>>> Using the same logic, we copy parents' employment and mothers' BMI as desired:
>>>
>>> . gen BMI_M = .
>>> (7 missing values generated)
>>>
>>> . gen Emp_M = .
>>> (7 missing values generated)
>>>
>>> . gen Emp_F = .
>>> (7 missing values generated)
>>>
>>> . summarize pid, meanonly
>>>
>>> . forval i = 1 / `r(max)' {
>>>   2.     by ID_fam: replace BMI_M = BMI[`i'] if ID_M == ID[`i'] & !missing(ID_M)
>>>   3.     by ID_fam: replace Emp_M = Emp[`i'] if ID_M == ID[`i'] & !missing(ID_M)
>>>   4.     by ID_fam: replace Emp_F = Emp[`i'] if ID_F == ID[`i'] & !missing(ID_F)
>>>   5. }
>>> (0 real changes made)
>>> (0 real changes made)
>>> (3 real changes made)
>>> (3 real changes made)
>>> (1 real change made)
>>> (0 real changes made)
>>> (0 real changes made)
>>> (0 real changes made)
>>> (0 real changes made)
>>> (0 real changes made)
>>> (0 real changes made)
>>> (0 real changes made)
>>>
>>>
>>> Here are the results:
>>>
>>> . l
>>>
>>>      +-----------------------------------------------------------------------------------------------+
>>>      |       ID_F         ID_M      BMI           ID     ID_fam   Emp
>>>  pid    BMI_M   Emp_M   Emp_F |
>>>      |-----------------------------------------------------------------------------------------------|
>>>   1. |                           26.501   A901963701   A9019637     1
>>>    1        .       .       . |
>>>   2. |                           20.483   A901963702   A9019637     1
>>>    2        .       .       . |
>>>   3. | A901963701   A901963702   20.924   A901963703   A9019637     .
>>>    3   20.483       1       1 |
>>>   4. |                           27.209   A901963801   A9019638     1
>>>    1        .       .       . |
>>>   5. |                           31.733   A901963802   A9019638     .
>>>    2        .       .       . |
>>>      |-----------------------------------------------------------------------------------------------|
>>>   6. | A901963801   A901963802   18.018   A901963803   A9019638     .
>>>    3   31.733       .       1 |
>>>   7. | A901963801   A901963802   19.054   A901963804   A9019638     .
>>>    4   31.733       .       1 |
>>>      +-----------------------------------------------------------------------------------------------+
>>>
>>> Nick
>>>
>>> On Fri, Feb 22, 2013 at 10:45 PM, Haena Lee <hannahlee419@gmail.com> wrote:
>>>
>>>> I am working on investigating the relationship between maternal
>>>> employment status and prevalence of childhood obesity using a
>>>> nationally representative data (KNHANES). Suppose I have ID(all
>>>> observations including both children and parents), ID_fam (household
>>>> indicator),
>>>> ID_F( father's ID), ID_M (mother's ID), BMI (body mass index) and
>>>> finally Emp (employment status 1 if employed; 0 if non-employed) as
>>>> the following;
>>>>
>>>> ID_F              ID_M           BMI                    ID                ID_fam       Emp
>>>>                                                  26.501         A901963701       A9019637   1
>>>>                                                  20.483         A901963702       A9019637   1
>>>> A901963701      A901963702       20.924         A901963703       A9019637    .
>>>>                                                  27.209         A901963801       A9019638   1
>>>>                                                  31.733         A901963802       A9019638    .
>>>> A901963801      A901963802      18.018            A901963803     A9019638    .
>>>> A901963801      A901963802      19.054          A901963804       A9019638    .
>>>>
>>>> And ultimately, I would like to have a data set like this following;
>>>>
>>>> ID (children)   ID_fam         BMI        Mom's Bmi Mom's Emp   Dad's Emp
>>>> A901963703  A9019637   20.924   20.483         1                    1
>>>> A901963803  A9019638   18.018   31.733          .                     1
>>>> A901963804  A9019638   19.054   31.733          .                     1
>>>>
>>>> Given this, my question is 1) how to map the properties of other
>>>> family members to children within each household, using loop, or 2)
>>>> how to generate an indicator of mother (1 if ID == ID_M; 0 otherwise)?
>>>> I found Nick Cox's helpful example and imitated it as the following;
>>>>
>>>> by ID_fam (ID), sort: gen pid = _n
>>>> gen byte fid = .
>>>> gen byte mid = .
>>>> summarize pid, meanonly
>>>> forval i = 1 / `r(max)' {
>>>>                 by ID_fam: replace fid = `i'
>>>>                 if ID_F == ID[`i'] & !missing(ID_F)
>>>>                 by ID_fam: replace mid = `i'
>>>>                 if ID_M == ID[`i'] & !missing(ID_M)
>>>> }
>>>>
>>>> And it didn't produce any meaningful values but missing. Please
>>>> advise. Thank you so much for any help in advance.
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/


© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index