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]

Does Varying spec limits by data category in variability chart give some ideas?

[#3]

Hi jthi,

I checked and followed steps, but it still didn't work, even directly used its attachment file and script. 

[#4]

Using your data (+ adding limits) and making small fix to the script from that thread

[#5]

Hi jthi,

You did indeed draw the chart I wanted, but I still don't know how to modify the script. Anyway, thanks for your reply

[#6 · Accepted solution]

Names Default To Here(1);

dt1 = Open("$DOWNLOADS/sample data.jmp");
dt = dt1 << Stack( // stack to get proper data format
	columns(:A10, :A20, :A30),
	Source Label Column("SampleGroup"),
	Output Table("Stack")
);

// Add limits as they were not provided
dt << New Column("LSL", Numeric, Continuous, Values([10, 10, 10, 10, 10, 10, 10, 10, 10, 20, 20, 20, 20, 20, 20, 20, 20, 20]));
dt << New Column("Target", Numeric, Continuous, Values([20, 20, 20, 20, 20, 20, 20, 20, 20, 40, 40, 40, 40, 40, 40, 40, 40, 40]));
dt << New Column("USL", Numeric, Continuous, Values([30, 30, 30, 30, 30, 30, 30, 30, 30, 50, 50, 50, 50, 50, 50, 50, 50, 50]));

var = dt << Variability Chart(Y(:Data), X(:Phase, :SampleGroup));

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

// We need access to the framebox
fb = Report(var)[Outline Box("Variability Chart for Data"), Frame Box(1)];

//Get phases count (this was missing)
Summarize(dt, phases = By(:Phase));

// Using graphic scripts most likely easiest option
xidx = 0;
For(i = 1, i <= N Items(phases), i++,
	curphase = dt_summary[i, "Phase"];
	curlsl = dt_summary[i, "LSL"];
	curusl = dt_summary[i, "USL"];
	curtarget = dt_summary[i, "Target"];
	curcount = dt_summary[i, "SampleGroup"];
	
	xend = xidx + curcount;
	
	lslstart = Eval List({xidx, curlsl});
	lslend = Eval List({xend, curlsl});

	uslstart = Eval List({xidx, curusl});
	uslend = Eval List({xend, curusl});

	targetstart = Eval List({xidx, curtarget});
	targetend = Eval List({xend, curtarget});
	
	xidx = xend;
	
	Eval(EvalExpr(
			fb << Add Graphics Script(
			Pen Color("Red");
			Pen Size(1);
			Line(Expr(lslstart), Expr(lslend)); // {x1, y1}, {x2, y2}
			Line(Expr(uslstart), Expr(uslend)); // {x1, y1}, {x2, y2}
			Pen Color("Blue");
			Line(Expr(targetstart), Expr(targetend)); // {x1, y1}, {x2, y2}
		);
	));
);

fb << Y Axis(Max(1.1*Max(Col Max(Column(dt, "Data")), Col Max(Column(dt_summary, "USL"))))); // adjust yaxis

[#7]

Hi jthi,

Thanks for your script, it's worked successfully. 

Regarding the limit setting below, if I have up to 100 data and only add limit line on dedicated "Phase", are there any solutions to optimize the following script? Thanks

// Add limits as they were not provided
dt << New Column("LSL", Numeric, Continuous, Values([10, 10, 10, 10, 10, 10, 10, 10, 10, 20, 20, 20, 20, 20, 20, 20, 20, 20]));
dt << New Column("Target", Numeric, Continuous, Values([20, 20, 20, 20, 20, 20, 20, 20, 20, 40, 40, 40, 40, 40, 40, 40, 40, 40]));
dt << New Column("USL", Numeric, Continuous, Values([30, 30, 30, 30, 30, 30, 30, 30, 30, 50, 50, 50, 50, 50, 50, 50, 50, 50]));

[#8]

You could create spec limit table or if you don't have too many parameters, associative array for that

Names Default To Here(1);

dt1 = Open("$DOWNLOADS/sample data.jmp");
dt = dt1 << Stack( // stack to get proper data format
	columns(:A10, :A20, :A30),
	Source Label Column("SampleGroup"),
	Output Table("Stack")
);

aa_specs = [
	"L1" => ["LSL" => 10, "USL" => 30, "Target" => 20],
	"L2" => ["LSL" => 30, "USL" => 50, "Target" => 40]
];

// Add limits as they were not provided
dt << New Column("LSL", Numeric, Continuous);
dt << New Column("Target", Numeric, Continuous);
dt << New Column("USL", Numeric, Continuous);

For Each Row(dt,
	:LSL = aa_specs[:Phase]["LSL"];
	:USL = aa_specs[:Phase]["USL"];
	:Target = aa_specs[:Phase]["Target"];
);

Or using spec table

Names Default To Here(1);

dt1 = Open("$DOWNLOADS/sample data.jmp");
dt = dt1 << Stack( // stack to get proper data format
	columns(:A10, :A20, :A30),
	Source Label Column("SampleGroup"),
	Output Table("Stack")
);

dt_specs = New Table("Specs",
	Add Rows(2),
	Compress File When Saved(1),
	New Column("Variable", Character, "Nominal", Set Values({"L1", "L2"})),
	New Column("LSL", Numeric, "Continuous", Format("Best", 12), Set Values([10, 30])),
	New Column("Target", Numeric, "Continuous", Format("Best", 12), Set Values([20, 40])),
	New Column("USL", Numeric, "Continuous", Format("Best", 12), Set Values([30, 50]))
);

dt << Update(
	With(dt_specs),
	Match Columns(:Phase = :Variable)
);

Close(dt_specs, no save);

[#9]

Hi jthi,

It worked when I used method1.

Assuming I want L2 to not have USL/LSL, how can I optimize it?

I tried setting value= 0, but a red line will show at the Y-zero position.

I put //"L2" => ["LSL" => 0, "USL" => 0, //"Target" => 40], it's not work.

aa_specs = [
	"L1" => ["LSL" => 10, "USL" => 30, "Target" => 20],
	"L2" => ["LSL" => 0, "USL" => 0, //"Target" => 40]
];

[#10]

I would set the value as missing if you don't wish to have it and I think it should work with the script. 

View more...

Names Default To Here(1);

dt1 = Open("$DOWNLOADS/sample data.jmp");
dt = dt1 << Stack( // stack to get proper data format
	columns(:A10, :A20, :A30),
	Source Label Column("SampleGroup"),
	Output Table("Stack")
);

aa_specs = [
	"L1" => ["LSL" => 10, "USL" => 30, "Target" => 20],
	"L2" => ["LSL" => 30, "USL" => ., "Target" => .]
];

// Add limits as they were not provided
dt << New Column("LSL", Numeric, Continuous);
dt << New Column("Target", Numeric, Continuous);
dt << New Column("USL", Numeric, Continuous);

For Each Row(dt,
	:LSL = aa_specs[:Phase]["LSL"];
	:USL = aa_specs[:Phase]["USL"];
	:Target = aa_specs[:Phase]["Target"];
);

var = dt << Variability Chart(Y(:Data), X(:Phase, :SampleGroup));

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

// We need access to the framebox
fb = Report(var)[Outline Box("Variability Chart for Data"), Frame Box(1)];

//Get phases count (this was missing)
Summarize(dt, phases = By(:Phase));

// Using graphic scripts most likely easiest option
xidx = 0;
For(i = 1, i <= N Items(phases), i++,
	curphase = dt_summary[i, "Phase"];
	curlsl = dt_summary[i, "LSL"];
	curusl = dt_summary[i, "USL"];
	curtarget = dt_summary[i, "Target"];
	curcount = dt_summary[i, "SampleGroup"];
	
	xend = xidx + curcount;
	
	lslstart = Eval List({xidx, curlsl});
	lslend = Eval List({xend, curlsl});

	uslstart = Eval List({xidx, curusl});
	uslend = Eval List({xend, curusl});

	targetstart = Eval List({xidx, curtarget});
	targetend = Eval List({xend, curtarget});
	
	xidx = xend;
	
	Eval(EvalExpr(
			fb << Add Graphics Script(
			Pen Color("Red");
			Pen Size(1);
			Line(Expr(lslstart), Expr(lslend)); // {x1, y1}, {x2, y2}
			Line(Expr(uslstart), Expr(uslend)); // {x1, y1}, {x2, y2}
			Pen Color("Blue");
			Line(Expr(targetstart), Expr(targetend)); // {x1, y1}, {x2, y2}
		);
	));
);

fb << Y Axis(Max(1.1*Max(Col Max(Column(dt, "Data")), Col Max(Column(dt_summary, "USL"))))); // adjust yaxis

You could also modify the script slightly and make sure it ignores those which have missing values, but it might not be necessary