[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
st: cutting panel data by date according to size
I have a panel data set of companies, with my time
variable "date" over many quarters, and my id variable
"co". I would like to categorise my data according to
size (my size variable is "TA"), into quintiles, BY
QUARTER. I want to get a single new variable, call it
quintile, that takes on values from 1-5 depending on
which quintile of size the company falls into in that
particular quarter. So for example a company might
have quintile=1 for a certain quarter in which it is
among the fifth of largest companies in that quarter,
and then quintile 2 for the next quarter if it is only
in the quintile of the second largest companies for
that quarter, and so on. So I would end up with an
equal number of companies in each of the 5 quintiles,
for each quarter. It is also important that I get a
single variable ("quintile") with values for every
company and every quarter (as opposed to separate
variables for each quarter).
I know this might sound complicated because of the
long-winded way I have explained it, but I think it
should be simple, I just don't know how to do it.
Doing it manually would be really horrible as I have
so many quarters. I thought that something like the
"cut" command under egen might do the trick. But the
problem is that "cut" cannot be used with the "by"
command., so it just gives 5 quintiles of equal size
over the whole sample, but not by quarter.
Any suggestions as to how to program this would be
Thank you! :)
Try the all-new Yahoo! Mail. "The New Version is radically easier to use" – The Wall Street Journal
* For searches and help try: