Thanks, Sebastian, I missed that point with the chosen coworkers.
Best,
Kristin
> Kristin,
>
> interesting idea but I do not think that it will work. Let say a 100
> subjects work in building in ten. Let also say subject no 7 is the
> subject one of interest and he/she also works in building ten. Now, five
> co-workers are chosen and three of them do also work in building ten.
> Then Carter would be interested in the mean of x for those three
> co-workers. Your approach, Kristin, will give the mean of all subjects
> working in building ten and not the mean of the co-workers enumerated in
> that list of max. five subjects in the same building.
>
> Regards
> Sebastian
>
> On 4/19/07, Kristin J. Kleinjans <[email protected]> wrote:
>> Hi Sebastian and Carter,
>> wouldn't
>> by bld: egen mean_x=mean(x)
>> do the same thing, or am I confused here?
>> Of course, it would include the x for the worker in question in the
>> mean, which might not be what you want or need.
>>
>> Regards,
>> Kristin J. Kleinjans
>>
>>
>>
>> > Carter,
>> >
>> > I do not think you need the -forvalues- or -foreach- command here.
>> What you are asking for can also be achieved with the following:
>> >
>> > //inputting example data
>> > clear
>> > input id bld bld1 bld2 bld3 bld4 bld5 x1 x2 x3 x4 x5
>> > 1 10 11 10 10 10 11 6 7 2 4 2
>> > 2 11 11 11 13 13 11 6 8 4 7 0
>> > 3 12 12 12 12 12 12 6 1 2 4 3
>> > 4 13 13 10 10 10 14 10 5 10 9 5
>> > 5 14 15 14 . . . 3 10 5 1 3
>> > 6 15 15 . . . . 2 6 5 2 7
>> > end
>> >
>> > // rename bld to bld_id
>> > ren bld bld_id
>> >
>> > // reshape the data to long structure
>> > reshape long bld x, i(id) j(cow_no 1 2 3 4 5)
>> >
>> > // determine the mean value of x for those coworkers within the same
>> // building - will not record anything if bld_id id missing (.)
>> > bys id: egen bld_mean_x = mean(x) if bld == bld_id & bld_id != .
>> >
>> > // make sure that values is entered in any observation belonging to
>> // an id
>> > bys id (bld_mean_x): replace bld_mean_x = bld_mean_x[1]
>> >
>> > // reshape the data back to wide
>> > reshape wide bld x, i(id) j(cow_no)
>> >
>> > The mean value for x for co-workers from the same building can now
>> be found in the variable bld_mean_x.
>> >
>> > Regards
>> > Sebastian
>> >
>> >
>> > On 4/19/07, Carter Rees <[email protected]> wrote:
>> >> Statalist,
>> >>
>> >> I have the variables id (person id), bld (person's building code),
>> bld1-bld5 (building codes for 5 nominated co-workers). I also have
>> various co-worker characteristics in my file. For illustration,
>> let x1-x5 be one of the characteristics listed for each co-worker.
>> The data are constructed as:
>> >>
>> >> id bld bld1 bld2 bld3 bld4 bld5 x1 x2
>> >> x3 x4 x5
>> >> 1 10 11 10 10 10 11 6 7
>> >> 2 4 2
>> >> 2 11 11 11 13 13 11 6 8
>> >> 4 7 0
>> >> 3 12 12 12 12 12 12 6 1
>> >> 2 4 3
>> >> 4 13 13 10 10 10 14 10 5
>> >> 10 9 5
>> >> 5 14 15 14 3 10
>> >> 5 1 3
>> >> 6 15 15 2 6
>> >> 5 2 7
>> >>
>> >> For each person id, I would like to construct a variable that is
>> the mean value of only those co-workers who are in the same
>> building. For example, the first case would only include
>> co-workers in building 10 which would result in an average of
>> (7+2+4)/3 = 4.33. -forvalues- and/or -foreach- seems appropriate
>> here but I can't seem to get it quite right.
>> >>
>> >> Many thanks,
>> >>
>> >> Carter
>> >>
>> >> *
>> >> * 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/
>>
>>
>>
>> *
>> * 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/
*
* 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/
