>> I have no idea how to "group-by" in the formula editor.
I didn't either, but by chance this came up last week in the forums. Took me a while to remember where it was (link). I'll post a bit of detail on it but there's a better way, I think.
If you open up the big class.jmp sample data file (jsl code open( "$SAMPLE_DATA\Big Class.jmp"); will open it), then add a column with the following formula, you'll see it contains the average height for each age group.
Formula is: Col Mean(:height, :age). But this is in a new column and won't show up in your variability gauge plot for whatever column (e.g. height) that you're plotting.
I think you might want to skip the formula step, summarize your column(s) by whichever group variable (table -> summary platform, put :mean(height) in stats box, :age in group box, and set stats name column format to 'column' for easy stacking later.
right click green triangle (or red in earlier versions) to edit and view source script:
Data Table( "Big Class" ) << Summary(
Group( :age ),
Mean( :height ),
Freq( "None" ),
Weight( "None" ),
statistics column name format( "column" )
)
Stack it
Data Table( "Big Class" ) << Concatenate( Data Table( "Big Class By (age)" ) )
Label the new rows. Label the :height column. Then do your plot. Should work ok.
You can change the marker from a black dot to something less like a data point (colored hozizontal line) so it is clear that these are not actual data points. (Edit: Exclude, but don't hide, the new rows to keep them from disturbing the valid data points.)
As far as I know there is no easy way to add text to plots.
