BookmarkSubscribeRSS Feed

Scripting a Gauge R&R Study

paigemiller

Community Trekker

Joined:

Apr 12, 2012

JMP 9

 

I would like to script a Gauge R&R study involving multiple variables, instead of using the red triangle dropdown menus multiple times from the Chart->Variabliity/Gauge Chart platform.

 

If you perform the analysis manually, and select from the red triangle dropdown menu item Gauge Studies/Gauge RR, you get two dialog boxes, one which asks if the design is crossed or nested (mine is crossed), and a second dialog box which asks for the Gauge R&R specifications.

 

When I create a script to perform this task over all Y variables, the first dialog box doesn't appear, but I cannot figure out how to get the 2nd dialog box to not appear, it always appears no matter what I do. If you know how to change the script so that the 2nd dialog box no longer appears, please let me know.

 

Here is my script:

 

 

rezults = Column Dialog( 
        title( "Select Columns For Gauge R&R Analysis" ), 
        yycol = ColList( "Y, Response",  Mincol( 1 ), DataType( Numeric ) ), 
        xxcol = ColList( "X, Factor",  Mincol( 1 ), Modeling Type( Nominal ) ), 
        bbycol = ColList( "By"  ));
ylist = rezults["yycol"]; 
xlist = rezults["xxcol"]; 
bbylist = rezults["bbycol"];
Variability Chart(
      Y( eval(ylist) ),
      X( eval(xlist) ),
      By ( eval(bbylist) ),
      Model( "Crossed" ),
      Max Iter( 100 ),
      Conv Limit( 0.00000001 ),
      Number Integration Abscissas( 128 ),
      Number Function Evals( 65536 ),
      Analysis Type( Name( "Choose best analysis (EMS REML Bayesian)" ) ),
      Historical Mean( 0 ),
      Historical Sigma( 1 ),
      Process Variation( 0 ),
      Std Dev Chart( 0 ),
      Gauge RR( 6, 0, 0, 0 ),
      Gauge RR Report( 1 )
);

 

2 REPLIES
mpb

Super User

Joined:

Jun 23, 2011

Something like the following might be helpful. I pulled up the sample dataset "2 factors crossed" and added a column I called MyY with formula random normal mean 0 var 1. The code below stacks the output in one window for all Y columns selected. If you omit the New Window wrapper each Y column would have its own window. The assumption is that each allowable column has values entered for the "Spec Limits" property.

 

 

nw = new window("Test Window",
 
rezults = Column Dialog(
        title( "Select Columns For Gauge R&R Analysis" ),
        yycol = ColList( "Y, Response",  Mincol( 1 ), DataType( Numeric ) ),
        xxcol = ColList( "X, Factor",  Mincol( 1 ), Modeling Type( Nominal ) ),
        bbycol = ColList( "By"  ));
 
ylist = rezults["yycol"];
xlist = rezults["xxcol"];
bbylist = rezults["bbycol"];
 
N = NItems(ylist);
 
//show(ylist,xlist,bbylist);
//show(ylist[1],xlist[1]);
 
for(i=1,i<=N,i++,
 
    specs = ylist << get property("Spec Limits");
    usl = specs["USL"];
    lsl = specs["LSL"];
 
    //show(specs);
    //show(usl);
    //show(lsl);
 
//)
 
    vc = Variability Chart(
        Y( ylist),
        X( :Operator, :part# ),
        Model( "Crossed" ),
        Max Iter( 100 ),
        Conv Limit( 0.00000001 ),
        Number Integration Abscissas( 128 ),
        Number Function Evals( 65536 ),
        Analysis Type( Name( "Choose best analysis (EMS REML Bayesian)" ) ),
        Historical Sigma( 0 ),
        Std Dev Chart( 1 ),
        Gauge RR( 6, usl-lsl, lsl, usl ),
        Gauge RR Report( 1 )
    );
 
)
)

 

Fiona0723

New Contributor

Joined:

May 16, 2018

Variability Chart only show 1st item on 5 kigma.
I want to get all item on 5 sigma. who can give me suggestion? 

 

rezults = Column Dialog(
title( "Select Columns For Gauge R&R Analysis" ),
yycol = ColList( "Y, Response", Mincol( 1 ), DataType( Numeric ) ),
xxcol = ColList( "X, Factor", Mincol( 1 ), Modeling Type( Nominal ) ),
bbycol = ColList( "By" )
);
ylist = rezults["yycol"];
xlist = rezults["xxcol"];
bbylist = rezults["bbycol"];
N = N Items( ylist );
For( i = 1, i <= N, i++,
s = dt << Get Property( "Spec Limits" );
usl = s["USL"];
lsl = s["LSL"];
);
obj = Variability Chart(
Y( Eval( ylist ) ),
X( Eval( xlist ) ),
By( Eval( bbylist ) ),
Model( "Main Effect" ),
Max Iter( 100 ),
Conv Limit( 0.00000001 ),
Number Integration Abscissas( 128 ),
Number Function Evals( 65536 ),
Analysis Type( "Choose best analysis (EMS REML)" ),
Historical Mean( 0 ),
Historical Sigma( 0 ),
//Process Variation( 0 ),
Std Dev Chart( 0 ),
);
obj << Gauge RR( 5, usl - lsl, lsl, usl );