How do I define spec limits by data category in variability chart

[#1]

Hi,

How can I draw spec limits(red line) by data category in variability chart as below example?  The red line was done by manual in pptx.

Attach jmp file data for reference. Appreciate for your reply. Thanks.

[#2]

Refer to attachments for test data and script.

[#3]

Like I said earlier, easiest option is to create new grouping column and use that in your Summary table. It might also be enough to just get sum of the n categories over multiple groups

curcount = Sum(dt_summary[i, {"Leg", "Layer"}]);

[#4]

Hi Jarmo

When I key multiple data, the spec limit does not follow script content. How could I modify script ?

[#5 · Accepted solution]

You have to calculate the group counts in different manner. One quite simple option could be to add new column to your dt which combines leg and layer

dt << new column("LEGLAYER", Character, Nominal, Formula(:Leg ||:Layer));

then use that in the summary

dt_summary = dt << Summary(
	Group(:SampleGroup),
	Max(:LSL),
	//Mean(:Target),
	Min(:USL),
	N Categories(:LEGLAYER),
	Freq("None"),
	Weight("None"),
	statistics column name format("column"),
	Link to original data table(0)
);

and get the curcount from that

	curcount = Sum(dt_summary[i, {"LEGLAYER"}]);

[#6]

Hi there,
If you're looking for a scripting-free solution and okay with using Graph Builder, it actually handles multi-level spec limits by category  — just make sure your spec limits are structured accordingly (e.g., via column property).

If you specifically want to stay with Variability Chart, this behavior isn’t built-in in core JMP — but it is supported in YieldOptiX, a plugin developed for test data analysis. It adds support for category-based spec limits directly in the Variability Chart.

So both options (Graph Builder or Variability Chart with grouped limits) are supported depending on your preferred interface — just depends on what fits your process best.