nav[aria-label="Primary Navigation"] { padding: 0; & ul { list-style: none; width: 100%; display: flex; flex-direction: row; justify-content: start; align-items: start; gap: 30px; padding: 0; & li { margin: 0; } & ul li { list-style: none; } } }

How to get Repeatability, Reproducibility and Gauge R&R Variation values for each parameter in a table post Gauge R&R analysis via JSL?

Neo

I am conducting Gauge R&R analysis for several (~100) measured parameters. My data table with parts, operatorID/Inspector and repeats has spec limits values saved into the measured parameters columns.

Interactively doing the Gauge R&R analysis process as shown at time 1.38 mins as in this video is inefficient and incontinent when the parameters are ~100.

https://community.jmp.com/t5/Statistical-Thinking-for/Conducting-a-Gauge-R-amp-R-Analysis/ta-p/271895https://community.jmp.com/t5/Statistical-Thinking-for/Conducting-a-Gauge-R-amp-R-Analysis/ta-p/271895

I would like to automate the process and get the the Repeatability, Reproducibility and Gauge R&R Variation values in a data table.

How to do this via JSL?

Accepted answers

jthi

I'm not sure which scripts you are currently talking about. This should use spec limits from column properties (except for the possible bug I did mention earlier which might be present in JMP16)

Names Default To Here(1);

dt = Open("$SAMPLE_DATA/Semiconductor Capability.jmp");

var = dt << Variability Chart(
	Y(:NPN1, :PNP1),
	Model("Main Effect"),
	X(:SITE),
	Historical Sigma(0)
);
var << "Gauge R&R"n();

If(N Items(var) > 1,
	result_ref = Report(var[1])[OutlineBox("Gauge R&R")];
,
	result_ref = Report(var)[OutlineBox("Gauge R&R")];
);

dt_result = result_ref[TableBox(1)] << Make Combined Data Table;

show(Column(dt, "NPN1") << Get Property("Spec Limits"));
show(Column(dt, "PNP1") << Get Property("Spec Limits"));

All comments

jthi

There are quite a many ways of doing this, below is fairly simple example of one possible way

Names Default To Here(1);

dt = Open("$SAMPLE_DATA/Semiconductor Capability.jmp");

var = dt << Variability Chart(
	Y(:NPN1, :PNP1, PNP2),
	Model("Main Effect"),
	X(:SITE),
	Variability Analysis(:NPN1, "Gauge R&R Report"n(1)),
	Variability Analysis(:PNP1, "Gauge R&R Report"n(1))	
);

if you have one-sided specifications be sure to check that the variables before and after those work correctly. There was a bug at some JMP version which did use wrong limits if you had one-sided somewhere in between the execution and you relied on spec limits column property (this might be fixed in JMP17, not sure. I use different method nowadays).

Neo

@jthi Thanks, I have both LSL and USL for each measured parameter.

What you have posted is, if I understand correctly, the initial analysis part.

I want to automate the next stage which is getting the Repeatability, Reproducibility and Gauge R&R Variation values (marked with red arrows) for each measured in a new data table.

In the video I linked before, this is done post 1.38 mins into the video and is done manually.

How to automate this part via JSL?

jthi

You right click on the Gauge R&R table and create a table from it by using Make Combined Data Table

From the resulting table's table script you can get an idea how to automate it (get a reference to table box and use Make Combined Data Table message)

Start automating with only few columns and when you get it working expand to more.

Neo

@jthi In this code, how to correctly loop through all the measured variables (~100) in my data table?

Names Default To Here(1);

dt = Open("$SAMPLE_DATA/Semiconductor Capability.jmp");

var = dt << Variability Chart(
	Y(:NPN1, :PNP1, PNP2),
	Model("Main Effect"),
	X(:SITE),
	Variability Analysis(:NPN1, "Gauge R&R Report"n(1)),
	Variability Analysis(:PNP1, "Gauge R&R Report"n(1))	
);

jthi

Build list of columns of interest and then send Gauge R&R Report message

Names Default To Here(1);

dt = Open("$SAMPLE_DATA/Semiconductor Capability.jmp");

var = dt << Variability Chart(
	Y(:NPN1, :PNP1),
	Model("Main Effect"),
	X(:SITE)
);

var << "Gauge R&R Report"n(1);

Neo

@jthi Thanks, This script partly does what I need to do, but needs me to press OK for each parameter's Gauge R&R analysis (I am on JMP 16.2) at the "Enter/Verify Gauge R&R specification" window. I do not want to do this for my ~100 parameters. How to automate the Gauge R&R analysis/ the press OK step?

Names Default To Here(1);

dt = Open("$SAMPLE_DATA/Semiconductor Capability.jmp");

var = dt << Variability Chart(
	Y(:NPN1, :PNP1),
	Model("Main Effect"),
	X(:SITE)
);

var << "Gauge R&R Report"n(1);

Also, I do not see the JMP generated script (when the OK button is pressed) in the log otherwise I could have got a direction to proceed.

Neo

@jthi Thanks. While I wait for understand how to automate this "OK" step for each measured parameter

If I try the following from above in my JM 16.2.0,

Names Default To Here(1);

dt = Open("$SAMPLE_DATA/Semiconductor Capability.jmp");

var = dt << Variability Chart(
	Y(:NPN1, :PNP1, PNP2),
	Model("Main Effect"),
	X(:SITE),
	Variability Analysis(:NPN1, "Gauge R&R Report"n(1)),
	Variability Analysis(:PNP1, "Gauge R&R Report"n(1))	
);
Wait (0);

rpt ["Variability Gauge", "Variability Gauge Analysis for PNP1", "Gauge R&R", Table Box[1]] <<Make Combined Data Table;
rpt << Close Window;

I get an error

Name Unresolved: rpt in access or evaluation of 'rpt' , rpt/*###*/

most likely as I haven't defined rpt. The following with script copied from JMP also throws an error

Names Default To Here(1);

dt = Open("$SAMPLE_DATA/Semiconductor Capability.jmp");

platform = dt << Variability Chart(
	Y(:NPN1, :PNP1, PNP2),
	Model("Main Effect"),
	X(:SITE),
	Variability Analysis(:NPN1, "Gauge R&R Report"n(1)),
	Variability Analysis(:PNP1, "Gauge R&R Report"n(1))	
);
Wait (0);

//rpt ["Variability Gauge", "Variability Gauge Analysis for PNP1", "Gauge R&R", Table Box[1]] <<Make Combined Data Table;
//rpt << Close Window;

Report( platform[1] )[Outline Box( "Gauge R&R" )][Table Box( 1 )] <<
Make Combined Data Table;
Report( platform[1] ) << Close Window;

I need some help to proceed.