I am trying to achieve a box plot graph where the confidence diamond and data points are shown in black, with the boxes itself using the color based on a column value. Additionally, I would like to show the raw data points with a jitter.
I am using a work-around suggested to me in a previous post:
How-to-change-all-box-plot-confidence-diamonds-colors-in-graph
That solution uses two multiple box plots on top of each other to allow for a different color for the box plot and its confidence diamond.
It mostly works, but the problem I still face is that potential outliers are now showing as two separate data points.
an example using big class.jmp
the code for this graph:
Graph Builder(
Size( 1845, 897 ),
Show Control Panel( 0 ),
Variables( Y( :weight ), Group X( :sex ), Color( :sex ) ),
Elements(
Box Plot( Y, Legend( 7 ), Box Style( "Solid" ), Fences( 0 ) ),
Points(
Y,
Color( 0 ),
Legend( 6 ),
Jitter( "Random Uniform" ),
Jitter Limit( 0.3179 )
),
Box Plot( Y, Color( 0 ), Legend( 8 ), Confidence Diamond( 1 ) ),
Bar( Y, Legend( 9 ), Bar Style( "Float" ), Label( "Label by Value" ) )
),
SendToReport(
Dispatch( {}, "400", LegendBox,
{Legend Position(
{7, [-1, 0, 1, -3], 6, [-1], 8, [-3, -3], 9, [-3, -3]}
)}
)
)
);
Is there a way to achieve this? I suppose it requires not drawing the box plot potential outlier, but I can imagine this is hard to do.