Reputation: 39
I'm doing a research about SATs scores by cities. I have every city coded. Is there any way to ask Stata to calculate the mean for each city without having to type the city-code? Not having to write: sum if city==1
and then sum if city==2
, but instead that Stata takes all the people with the same city code and calculates the mean?
Upvotes: 0
Views: 910
Reputation: 9460
There are several ways to skin a cat, but something like this would do it:
table city_name, contents(mean sat_score count sat_score) row
This also gives you the number of individuals with non-missing scores in each city as well as an overall average. Type help table
for other options.
sum
also takes a bysort
prefix, so bysort city_name: sum sat_score
also works, but the output is less compact than with table
.
Unlike the egen
approach, neither of these will create a variable that stores the city-specific means. It will only show you the output on the screen.
Upvotes: 1
Reputation: 151
It looks like this post on stack overflow has some code it in you could use:
Using if qualifier with egen in Stata
Upvotes: 1