Running Maximum by group

[Costas Antoniou]

Hello Everyone

I have a dataset that has observations on an X variable for different individuals over time. I would like to calculate new variables that equal the maximum and minimum of X for each individual when considering all the values up to time t.

So for example

Individual time X max min
1 1 1 . .
1 2 -1 1 -1
1 3 3 3 -1

Thank you.

Costas

[Nick Cox]

https://www.stata.com/support/faqs/d...m-of-sequence/

The FAQ explains how to do it with official commands only, but my bias is to plugging rangestat as getting both variables in one line:

clear
input id time X
1 1 1
1 2 -1
1 3 3
end

ssc install rangestat

rangestat (max) max=X (min) min=X, int(time . 0) by(id)

list

     +----------------------------+
     | id   time    X   max   min |
     |----------------------------|
  1. |  1      1    1     1     1 |
  2. |  1      2   -1     1    -1 |
  3. |  1      3    3     3    -1 |
     +----------------------------+

[Costas Antoniou]

Thank you Nick, this worked perfectly. My date variable was an actual date, and so to simplify things and avoid using interval() I sorted by id and date and created a counter (by id: g time=_n).

[Nick Cox]

That shouldn't be needed with any numeric date variable. It's exactly the same logic for interval(date . 0)