'collapse' data

[J McClure]

Hi,
I have data from a one month period of time in an emergency room. The
data is by diagnosis not by individuals (var is Dx). For each of
diagnosis of 58 diagnoses I have the number of times it was listed as
the diagnosis (var 'total') for 224 individuals admitted to the
emergency department for that month. Because the 'total'  reflects
listing of a diagnosis as a primary or secondary one, the  'total' sums
to more than 224.  I created a new variable (DxGrp) and assigned each of
the 58 diagnoses to one of the 7 diagnostic groups. I want to 'collapse'
the information such that I have 7 diagnostic groups and the number of
times that group was the primary or secondary diagnoses.
I tried aggregate but got the number of times each DxGrp was found in
the list of 58 diagnose. I considered count but I don't think I need to
sum values across variables. I also considered compute but can't figure
out how to write the syntax.
Any help appreciated. Jan

Here is a sample of the data.
Dx POSTTRAUMATIC STRESS DIS   total 55   DxGrp 5
Dx DEPRESSIVE DISORDER NEC       total 39   DxGrp 1
Dx ALCOH DEP NEC/NOS-UNSPEC    total 16    DxGrp 7
Dx RECURR DEPR DISORDER-MOD  total 14    DxGrp 1

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[J McClure]

Thanks! Your code worked.
Jan

On 9/2/2010 10:13 AM, Marks, Jim wrote:

> Can you post your aggregate syntax?
>
> Something like this should have worked:
>
> DATASET DECLARE sumfile.
> AGGREGATE OUTFILE = sumfile /BREAK = diaggrp /cases = SUM (total).
> DATASET ACTIVATE sumfile'
>
>
> Jim Marks
> Director, Market Research
> x1616
>
>
> -----Original Message-----
> From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
> J McClure
> Sent: Thursday, September 02, 2010 12:07 PM
> To: [hidden email]
> Subject: 'collapse' data
>
> Hi,
> I have data from a one month period of time in an emergency room. The
> data is by diagnosis not by individuals (var is Dx). For each of
> diagnosis of 58 diagnoses I have the number of times it was listed as
> the diagnosis (var 'total') for 224 individuals admitted to the
> emergency department for that month. Because the 'total'  reflects
> listing of a diagnosis as a primary or secondary one, the  'total' sums
> to more than 224.  I created a new variable (DxGrp) and assigned each of
> the 58 diagnoses to one of the 7 diagnostic groups. I want to 'collapse'
> the information such that I have 7 diagnostic groups and the number of
> times that group was the primary or secondary diagnoses.
> I tried aggregate but got the number of times each DxGrp was found in
> the list of 58 diagnose. I considered count but I don't think I need to
> sum values across variables. I also considered compute but can't figure
> out how to write the syntax.
> Any help appreciated. Jan
>
> Here is a sample of the data.
> Dx POSTTRAUMATIC STRESS DIS   total 55   DxGrp 5
> Dx DEPRESSIVE DISORDER NEC       total 39   DxGrp 1
> Dx ALCOH DEP NEC/NOS-UNSPEC    total 16    DxGrp 7
> Dx RECURR DEPR DISORDER-MOD  total 14    DxGrp 1
>
> =====================
> To manage your subscription to SPSSX-L, send a message to
> [hidden email] (not to SPSSX-L), with no body text except the
> command. To leave the list, send the command
> SIGNOFF SPSSX-L
> For a list of commands to manage subscriptions, send the command
> INFO REFCARD
>
>

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD