# st: RE: difficult merging issue

 From "Nick Cox" To Subject st: RE: difficult merging issue Date Sun, 13 Jul 2008 16:48:45 +0100

```I do not understand this to be a merging issue. You appear to have just
one dataset.
In Stata, -merge- requires two datasets.

Nor do I see why you think you need a dummy for every official.

My guess is that you have a problem for which -collapse- is the
solution.

However, you need to -reshape- first.

Here is a guess at an analogue to your problem.

. l

+--------------------------------------+
| Meet   WhoL   ScoreL   WhoW   ScoreW |
|--------------------------------------|
1. |    1      B        3      A       12 |
2. |    2      A        6      B        7 |
3. |    3      D        3      C        4 |
4. |    4      C        4      D        3 |
+--------------------------------------+

. reshape long Who Score, i(Meet) string
(note: j = L W)

Data                               wide   ->   long
------------------------------------------------------------------------
-----
Number of obs.                        4   ->       8
Number of variables                   5   ->       4
j variable (2 values)                     ->   _j
xij variables:
WhoL WhoW   ->   Who
ScoreL ScoreW   ->   Score
------------------------------------------------------------------------
-----

. l

+-------------------------+
| Meet   _j   Who   Score |
|-------------------------|
1. |    1    L     B       3 |
2. |    1    W     A      12 |
3. |    2    L     A       6 |
4. |    2    W     B       7 |
5. |    3    L     D       3 |
|-------------------------|
6. |    3    W     C       4 |
7. |    4    L     C       4 |
8. |    4    W     D       3 |
+-------------------------+

After that, you can -collapse- by -Who-.

Nick
n.j.cox@durham.ac.uk

Rufus Peabody

I'm doing an analysis of college football officiating and have a large
dataset.  Each game played is an observation.  There are 7 types of
officials, and I have an ID for each official.  I have the 7 official
variables, but the problem is that officials are not confined
exclusively to one type---for example, one official might be "Referee"
one game and "Line Judge" another.  What I did to alleviate this
problem was create dummies for each official (there are 596 in total).

Now, the major issue:  I would like to somehow transform this so that
each official is an observation and I have mean(penalties),
mean(penalty yards), total games, etc. as variables.  This would mean
there should be 596 total observations.  Anybody have any suggestions
of how to tackle this problem?

*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/
```