Adding Spec limits in variability-attribute gauge chart

Vmuthu

I have a table with 100 devices, and 300 test items tested. I have spec columns separately contains LSL, USL. I want to plot the variability-attribute gauge chart with the spec limits in it. Please show me how to do it or if there is any existing discussion in this topic, please point it to me. thanks.

Sample data table:

Device	Test Name	Measurement	LSL	USL
1	Test1	230	200	240
1	Test2	20	15	25
1	Test3	30	25	35
2	Test1	234	200	240
2	Test2	22	15	25
2	Test3	29	25	35
3	Test1	229	200	240
3	Test2	21	15	25
4	Test1	230	200	240
4	Test2	21	15	25

 

 

txnelson

The main issue with what you want, is that your data are not in a proper form for running the Variability-Attribute Chart platform.  The script below will transform a data table with the columns you have specified, into a data table that can be used to generate the charts you want.

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

// Make a list of the Tests
Summarize( dt, testList = by( :Test Name ) );

dtSplit = dt << Split(
	Split By( :Test Name ),
	Split( :Measurement, :LSL, :USL ),
	Group( :Device ),
	Sort by Column Property
);

// Rename the test columns and set the Spec Limits
For( i = 2, i <= N Items( testList ) + 1, i++,
 
	// Assign the new names
	Column( dtSplit, i ) << set name( Substr( Column( dtSplit, i ) << get Name, 13 ) );
	
	// Set the Spec Limits
	Eval(
		Substitute(
				Expr(
					Column( dtSplit, i ) << set property(
						"Spec Limits ",
						{LSL( __LSL__ ), USL( __USL__ ), Show Limits( 1 )}
					)
				),
			Expr( __LSL__ ), Col Mean( As Column( dtSplit, i + N Items( (testList) ) ) ),
			Expr( __USL__ ),
				Col Mean( As Column( dtSplit, i + 2 * N Items( (testList) ) ) )
		)
	);
);

// Delete unwanted limits columns
dtSplit << delete columns( Index( N Items( testList ) + 2, N Cols( dtSplit ) ) );

 

example.jmp