How to calculate sums across duplicate rows in SPSS 11.5

[sheree]

I have a dataset in SPSS that looks like this:

StudyID Time Temp DegMins
01-35-001 1 36.9 13.5
01-35-001 2 36.2 3
01-35-001 3 34.9 -16.5
01-35-001 4 35.6 -6
01-35-001 5 35.5 -7.5
01-35-002 1 36.5 7.5
01-35-002 2 36.5 7.5
01-35-002 3 36.4 6
01-35-002 4 35.8 -3
01-35-003 1 36.9 13.5
01-35-003 2 36.8 12
01-35-003 3 36.7 10.5

Where the StudyID will repeat across several rows -- this is the same
person, with measurements at different times. What I want to do, is
create a new variable that sums the DegMins values **for each
StudyID**, but only when the DegMins are a negative number. For
example, for StudyID=01-35-001, Sum=(-16.5)+(-6)+(-7.5)=-30. For
StudyID=01-35-002, Sum=(-3)=-3. For StudyID=01-35-003, Sum=0 (since
there are no negative values for this person). Notice also that the
number of rows that exist for each StudyID will vary.

Ideally, I would like to know the SPSS syntax for how to do this, but
any way to get this answer & store it would be great.

Thank you very much for your suggestions!
-sheree

[Moshe Braner]

compute negonly = max(degmins,0).
aggregate outfile=* /break=studyid /sumneg=sum(negonly).

[Rich Ulrich]

You don't have to save the Recode, so you just don't
even need to set up a temporary variable to use. Just -

recode DegMins(0 thru hi=0).

- and Aggregate, by StudyID, saving the new file
which computes a SUM as its variable.
That gives you one record for each StudyID, with
the sum of whatever was negative.

Right?

--
Rich Ulrich, wpi...@pitt.edu
http://www.pitt.edu/~wpilib/index.html
"Taxes are the price we pay for civilization."