-egen, group()- maps all combinations
that occur in your data to integers 1 up.
If you want all possible combinations
to be coded systematically, you need
some binary representation. Assuming
that -var1- to -var5- are coded 0 or
1 this could be
gen comb = 16 * var1 + 8 * var2 + 4 * var3 + 2 * var4 + var5
or something producing a corresponding string.
Often the latter is more transparent. There is a
corresponding -egen- function -base()- in the
package -egenmore- on SSC.
Nick
[email protected]
riteshm
>
> I need help with a data management problem. What is an
> efficient way of
> creating a categorical variable which consists of all unique
> combinations of a series of indicators variables?
>
> For example, I have a data set with five binary variables: var1, var2,
> var3, var4 and var5. A new variable is needed in which the categories
> represent all possible unique combinations of var1-var5.
*
* 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/
