JSL: Profiler Remember Settings

Marco_

Hi everyone,

 

I have a model in the profiler and would like to save a bunch of different settings. Usually, I use the Remember Parameter Setting in the Profiler to do this. However, in this particular case I want to save all combinations for the parameters X1 to X4 from -1 to +1 in 0.1 steps.

I need some help with the following JSL Code to do this:

Profiler(
	Y( :Y ),
	Profiler(
		1,
		Confidence Intervals( 1 ),
		Term Value(
			:"X1"n( 0, Lock( 0 ), Show( 1 ) ),
			:"X2"n( 0, Lock( 0 ), Show( 1 ) ),
			:"X3"n( 0, Lock( 0 ), Show( 1 ) ),
			:"X4"n( 0, Lock( 0 ), Show( 1 ) )
		),
		<FONT color="#FF0000">Remember Settings( "Setting 1", Differences Report( 1 ) ),
		Term Value(
			:"X1"n( 0, Lock( 0 ), Show( 1 ) ),
			:"X2"n( 0, Lock( 0 ), Show( 1 ) ),
			:"X3"n( 0, Lock( 0 ), Show( 1 ) ),
			:"X4"n( 0, Lock( 0 ), Show( 1 ) )
		)
			)</FONT>
		);

It would be awesome if someone could tweak the existing code.

 

Thanks in advance!

 

jthi

You should be able to use Full Factorial Design to create all possible combinations.

 

I would suggest NOT running this with the all possible combinations

Names Default To Here(1);
Clear Log();

// Calculate combinations - START
doe_platform = DOE(
	Full Factorial Design,
	{Add Response(Maximize, "Y", ., ., .), Add Factor(
		Continuous,
		{1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21},
		"X1",
		0
	), Add Factor(
		Continuous,
		{1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21},
		"X2",
		0
	), Add Factor(
		Continuous,
		{1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21},
		"X3",
		0
	), Add Factor(
		Continuous,
		{1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21},
		"X4",
		0
	), Set Random Seed(6613116), Make Design, Simulate Responses(0),
	Set Run Order(Sort Left to Right)},
	Invisible
);

dt = doe_platform << Make Table(Invisible);
doe_platform << Close Window;

m = dt[0, 2::5];
Close(dt, no save);
Show(N Items(m), Dim(m), N Row(m) == 21^4); // N Items(m) = 777924; Dim(m) = [194481 4]; N Row(m) == 21 ^ 4 = 1;

settings = As List((m - 11) / 10);

m = Empty();
// Calculate combinations - END


dt = Open("$SAMPLE_DATA/Tiretread.jmp", Invisible);
dt << New Column("A", Numeric, Continuous, Formula(
	:SILICA * :SILANE * :SULFUR * :ELONG
));

obj = dt << Contour Profiler(
	Y(A),
	Invisible
);

dt << Add Rows(1);
max_row = N Rows(dt);
For Each({setting}, settings[1::100],
	dt[max_row, {"SILICA", "SILANE", "SULFUR", "ELONG"}] = setting;
	obj << Set to Data in Row(max_row);
	obj << Remember Settings;
);
dt << Delete Rows(max_row);

obj << show window;
Write();