Event History Analysis

[Charles Ngiendo]

Hi,
I have a longitudinal data from 2004 to 2010.
I need to follow only individuals from the year 2004 onwards. The individuals have specific id's. Is there a specific code I could use to keep only the individuals who exists only in 2004 onwards and get off the rest? For instance new individuals entering the dataset from say 2005 onwards should NOT be excluded without deleting everything.

[William Lisowski]

Suppose you have a variable named year that takes the values 2004 through 2010 and a variable named ID that gives the individual's identifier.

by ID (year), sort: keep if year[1]==2004

will keep all the observations for only those individuals for which their first year in the data is 2004. Any individual who entered after 2004 will not be kept.

[Charles Ngiendo]

Thank you. I'll run this

[Charles Ngiendo]

Hi @William Lisowski, will this eliminate all the other variables in the process or simply keep the variables and eliminate the individuals entering from 2005 with their respective variables?

[William Lisowski]

You should read the output of

help keep

which will describe the keep and drop commands.

The short summary is that the keep and drop commands each operate on variables or on observations, but not on both in a single command.

I'm thinking perhaps you have used keep or drop to remove variables in the past without reading the documentation and understanding that they each serve two tasks.