Data Points added to Box Plots I think there is an opportunity to improve the box plot graphs by adding the data points from the series.

The layout is similar to that produced in JMP and I have found that the addition of the data points to the box plots very useful.

In the attached example I overlayed the box and whisker on top of the datapoints. For Series A, the label is set at X=0, Series B at x = 1, and Series C at x =2. I created a wave for the x values for each series with x = 0 + gnoise(.3), x = 1 + gnoise(.3).... and then plotted the series values versus the x value waves. The reason for the jitter is to show points that may overlap because they have the same value.

This adds another dimension to looking at the data. I think it would be an easy addition. Everything for the folks at Wavemetrics is easy. Update: The ability to do this sort of thing is present in 6.2. From the what's new file:

Added /NCAT flag to AppendToGraph. Causes trace to be plotted normally on what otherwise is a category plot. x values are just category numbers but can be fractional. Category numbers start from zero. This can be used to overlay the original data points for a box plot. Here are examples:

Make/T/O catx={"cat0", "cat1", "cat2"}
Make/O caty= { 1, 3, 2}
Display caty vs catx
SetAxis/A/E=1 left

// Example 1: plot a function vertically overlaying a category
Make/O cat0over= 0.5*(sin(p/8)+1)
SetScale x, 0,3,cat0over
AppendToGraph/NCAT/VERT cat0over

// Example 2: simulate original data for a box plot
Make/N=10/O cat1over= gnoise(1)+1.5
SetScale/P x, 1.5, 1e-5, cat1over   // delta x can not be zero
AppendToGraph/NCAT cat1over
ModifyGraph mode(cat1over)=3,marker(cat1over)=19,rgb(cat1over)=(0,0,65535)

Dear Dr. Hededus,

I am a very naive beginner of the use of Igor, and I found your fascinating post about Data points added to Box plots. I would really appreciate if you can teach me how to create this graph.