Creating Spec limits

[#1]

Hi, 

I'm trying to create spec limits using jmp interface or jsl. I'm trying to create the spec limit for different columns with no USL or LSL.. Please help. thank.

[#2]

Have you looking into the Limits Manager?

     Analyze=>Quality and Process=>Limits Manager

[#3]

Hi Jim, 

I have looked at it, but from what I see this tool is to accept limit and apply it to column. What I'm trying to do is to make just calculate a USL or LSL based on a sigma, and a mean that I provide. Or if there is a tool where I specific what I want the sigma to be for that column and then JMP generates the spec limits. Have you been across something like this ? thanks Jim.

[#4 · Accepted solution]

Spec Limits are not normally calculated limits.  They are set based on the design specifications or on the tool limitations, etc.

However, here is a simple script that will set the spec limits for all continuous columns in a data table using 3 sigma limits

Names Default To Here( 1 );
dt = Current Data Table();

colNames = dt << get column names( string, continuous );

For Each( {col}, colNames,
	theMean = Col Mean( As Column( dt, col ) );
	theSTD = Col Std Dev( As Column( dt, col ) );
	Eval(
		Substitute(
				Expr(
					Column( dt, col ) << set property(
						"Spec Limits",
						{LSL( _LSL_ ), Target( _Target_ ), USL( _USL_ ), Show Limits( 1 )}
					)
				),
			Expr( _LSL_ ), theMean - 3 * theSTD,
			Expr( _Target_ ), theMean,
			Expr( _USL_ ), theMean + 3 * theSTD
		)
	);
);

[#5]

is it possible to setup different spec limit for different step w/o splitting all steps into each column? 

[#6]

Currently, only one set of spec limits are allowed per column.

However, using JSL, multiple spec limits can be woven into a script that can loop through a data table pulling out each subgroup one at a time, applying the spec limits for the subset and then running the analysis.  Here is a simple example.

Names Default To Here( 1 );
dt = 
// Open Data Table: Big Class.jmp
// → Data Table( "Big Class" )
Open( "$SAMPLE_DATA/Big Class.jmp" );

dtLimits = New Table( "Limits",
	New Column( "Group", character, values( {"F", "M"} ) ),
	New Column( "USL", values( {66, 70} ) ),
	New Column( "Target", values( {61, 64} ) ),
	New Column( "LSL", values( {52, 51} ) )
);

groups = Associative Array( dt:sex ) << get keys;

nw = New Window( "Distributions", hlb = H List Box() );

For Each( {level}, groups,
	dt << select where( :sex == level );
	dtSub = dt << subset( selected columns( 0 ), selecte rows( 1 ), output table( level ), invisible );
	theRow = (dtLimits << get rows where( :group == level ))[1];
	Eval(
		Eval Expr(
			dtSub:height << set property(
				"spec limits",
				{LSL( Expr( dtLimits:LSL[theRow] ) ), USL( Expr( dtLimits:USL[theRow] ) ),
				Target( Expr( dtLimits:Target[theRow] ) ), Show Limits( 1 )}
			)
		)
	);
	hlb << append(
		dis = dtSub << Distribution( Continuous Distribution( Column( :height ), Process Capability( 0 ) ) );
		Report( dis )[Outline Box( 2 )] << set title( "Height=" || level )
		;
	);	
);

[#7]

Certainly! To create spec limits in JMP using JSL, open your data table, choose the desired column, and use the Distribution platform with the Spec Limits option to set your limits.