calculate mean by group with condition.

[River Huang]

Suppose that I have the following variable, hhid, age, and x.

* Example generated by -dataex-. To install: ssc install dataex
clear
input float(hhid age x mean_x mean_x50)
1 45 1 2.5 2.5
1 51 2 2.5 2.5
1 65 3 2.5 2.5
1 33 4 2.5 2.5
2 30 5 6.5 7.5
2 45 6 6.5 7.5
2 60 7 6.5 7.5
2 62 8 6.5 7.5
end

I calculate the mean by group (hhid) with

bys hhid: egen mean_x = mean(x)

and I also want to do that again, but only for age > 50. The desired results are shown as "mean_x50". Any suggestions? Thanks.

[Nick Cox]

bys hhid: egen mean_x50_1 = mean(x) if age > 50  
bys hhid: egen mean_x50_2 = mean(x / (age > 50))    
bys hhid: egen mean_x50_3 = mean(cond(age > 50, x, .))

The second and third have identical consequences, but the first only puts results in observations it uses.

See http://www.stata-journal.com/sjpdf.h...iclenum=dm0055 Sections 9 and 10.

[River Huang]

bys hhid: egen mean_x50_1 = mean(x) if age > 50
bys hhid: egen mean_x50_2 = mean(x / (age > 50))
bys hhid: egen mean_x50_3 = mean(cond(age > 50, x, .))

The second and third have identical consequences, but the first only puts results in observations it uses.

See http://www.stata-journal.com/sjpdf.h...iclenum=dm0055 Sections 9 and 10.

Nick, many thanks for the help.

[Ridwan Sheikh]

Hey Nick,
i have a following question
please help..

[Ridwan Sheikh]

I have 4-dimensional panel data of imports of country (j) from country (i) in sector (s) at time (t). The number of importing countries is 88 (j=1 to 88) and exporters is 24 (i = 1 to 24). In this data, i have a variable called ''distance_ij" which gives bilateral distance between (i) and (j) . I want to generate two variables , for example:
(1) mean distance of country (j) from all the countries in (i) in each sector .
(2) mean distance of country (i) from all the countries in (j) for each sector. i am running the following codes ;
egen x1 = mean(distance_ij), by (importer sector)
egen x2 = mean(distance_ij), by (exporter sector)
i want to know whether these are right codes to generate these mean distance.
in my case i am using Baier and Bergstrand, (2009) approach of gravity model .
please help..
thanks,
Ridwan