This is a lot clearer. In addition to direct attacks using -twoway-
explicitly,
separate Score, by(Task)
then
linkplot Score? Character, link(GRP)
but that would be a mess.
To get a slide show, assuming prior use of -separate- as above,
forval i = 1/203 {
linkplot Score? Character if GRP == `i', link(GRP) asyvars
/// subtitle(GRP `i')
more
}
Nick
[email protected]
Stephen Cox
Hmmm, sorry for my confusing first attempt. After I sent it I realised I
had not said I had changed my data to long form.
Here is my data format:
GRP Task Character Score
1 1 1 40
1 1 2 44
1 1 3 11
1 1 4 78
1 1 5 38
1 2 1 42
1 2 2 46
1 2 3 17
1 2 4 71
1 2 5 42
2 1 1 40
2 1 2 44
2 1 3 11
2 1 4 78
2 1 5 38
2 2 1 42
2 2 2 46
2 2 3 17
2 2 4 71
2 2 5 42
Etc
For each GRP (group), I am wanting to get two line graphs for Score
against Character, one graph for each Task. And I want all graphs
overlayed.
Lets assume I have only the values for Task = 1.
After defining the dat set to be a panel data using
. tsset GRP character
Then using the following command:
.xtline score, overlay legend (off)
I can get a graph for Task = 1 in the form I want.
But I want to have another graph for Task=2, and have those graphs
overlayed.
I am unsure how to do this with -xtline-, as I think I would need two
'score' variables, one for each Task, but then I can't use the overlay
option.
I am not sure how to do this with -linkplot- (from SSC) either.
And ultimately I will have to do it all for some different subsets of
cases as I have 203 GRPs which will be too many too see on one graph,
but that is another learning task for me.
*
* 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/
