I have the following data:
ID aaa bbb ccc ddd YEAR
with multiple years per ID (i.e. 1978-2004) and 50 different discrete IDs
(actually 1300 or so observations, then). As it is time series, I want to
examine the trends by year for each ID, so I want to plot line graphs for
each of the four variables (aaa bbb ccc ddd), by each ID. Obviously
pasting the synnatx for each split the file and variable combination is a
chore!!! As an aside, just splitting the file and running the graph
procedure for each variable (i.e. 4 procedures in this example) while
producing a nice array of line graphs is not useful as the graphs are
basically measured on the same scale, but in actuality the differences
invalues for the variables between the IDs are quite large, thus the IDS
with high variability swamp those with low variability.
It seems to me that a macro can be set up to do the analysis (line graph)
for the first variable (aaa) for the first group (ID = 1) and then move on
to the next procedure (probably would make sense to to aaa by ID=2 next).
Unfortunately I am not there yet for writing this procedure.
=====================
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
Locked
