Nurhan,
It is indeed possible to -collapse- by more than one variable. The
-by()- option accepts a list of variables.
. sysuse auto
. collapse mpg, by(rep78 foreign)
. clist, noobs
rep78 foreign mpg
1 Domestic 21
2 Domestic 19.125
3 Domestic 19
3 Foreign 23.3333
4 Domestic 18.4444
4 Foreign 24.8889
5 Domestic 32
5 Foreign 26.3333
. Domestic 23.25
. Foreign 14
In this example, rep78 stands for the household ID, foreign for the
type of expense, and mpg for the amount.
Another option would be to create new variables containing the
expenditures on food and clothing.
. sysuse auto, clear
. gen dom_mpg = mpg if foreign == 0
. gen for_mpg = mpg if foreign == 1
. collapse dom_mpg for_mpg, by(rep78)
. clist, noobs
rep78 dom_mpg for_mpg
1 21 .
2 19.125 .
3 19 23.33333
4 18.44444 24.88889
5 32 26.33333
. 23.25 14
rep78 stands for the household ID, dom_mpg for food expenses, and
for_mpg for clothing expenses. In this case the structure of the final
data set matches your example.
Friedrich
P.S. Do you have a version of Stata that accepts variable names that
do not begin with a letter or underscore?
On 7/18/07, dnurhan <[email protected]> wrote:
My data set looks like:
ID $AMOUNT TYPE
4001 $10 1=food
4001 $25 1=food
4001 $12 2=clothing
4001 $33 2=clothing
4002 .... ...
4002 .... ...
i would like to sum across both $AMOUNT & TYPE, to get each family's
spending on food and clothing as
ID FOOD CLOTHING
4001 $35 $45
4002 .... .....
How do i do this?? (i am new to the list, looked at the manuals but they
explain using "collapse" across only one
variable)
Best regards, nurhan
*
* 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/
