Removing Cases Under Threshold Value

[Chris Felts]

Hello,

I'm new to Stata so forgive my ignorance. I have searched the forums for this answer but cannot find it.

I am cleaning a data set with Stata. I have a categorical variable "cal". I want to remove any categories that have fewer than 15 observations. In other words, if 85 respondents selected "guitar" but only 4 respondents selected "flute", I'd want to remove all cases that responded "flute" but keep the cases that selected "guitar".

How do I do this?

Thanks in advance.

[Chris Felts]

I have tried the code:

drop if cal < 15

But that is obviously he variable value not the frequency within a category. How do I designate that the 15 refers to a count or frequency?

[William Lisowski]

Perhaps this will start you on a useful path.

by cal, sort: generate count = _N
drop if count<15

The by prefix causes the command to be run separately for each group of observations with a common value of cal. While _N is usually the number of observations in a dataset, with the by prefix it becomes the number of observations in the group.

[Chris Felts]

Took care of it. Thanks William!