Creating a new database from aggregate data

[KMack]

I am trying to make a new dataset based on output from another and I'm not sure even where to start.   Any direction is highly appreciated.

Here is a simplified version of what I'm trying to do.  I have a person-centered dataset, say with 100 people coming from a total of 10 communities.  I have a number of binary variables for each person (e.g. gender (m/f) and current smoker (y/n)).  I'm interested in creating community variables for each binary variable in these individual data, e.g. % male in each community and % smoker in each community.  Then I would like to use this community information in subsequent analyses (so now my n=10).

I'm seriously tempted to manually input the column % statistic from x*community frequency tables but I'm convinced there is a much better way, right?

[Tom · Accepted solution]

General PROC SUMMARY (proc means) is the easiest way to summarize the data.  For a binary (1/0) variables the mean is the same as the percent that have a 1.  So you could do something like this:

proc summary data=have  nway ;

  class community ;

  var sex smoker ;

  output out=want mean=male_percent smoker_percent ;

run;

[KMack]

Thanks!  Proc summary generated exactly the output I was looking for.

[Ksharp]

You'd better to post some sample data and what output you need.

It will give you answer fastestly and correctly.

Ksharp