BookmarkSubscribeRSS Feed
mkennke

Community Trekker

Joined:

Mar 18, 2016

jsl colorize profiler and show data in Marginal Model Plots

Hi,

 

can anybody tell me how to activate the "Colorize Profiler" (Variable Importance) and "Show Data" (Marginal Model Plot) function with JSL?

 

Following a test script where i can activate the coloration and the data by hand but not via JSL.

Names Default To Here( 1 );
dt = Open( "$SAMPLE_DATA/Tiretread.jmp" );
Profiler(
	Y( :Pred Formula ABRASION, :Pred Formula MODULUS, :Pred Formula ELONG, :Pred Formula HARDNESS ),

	Expand,
	Dependent Resampled Inputs( 1 ),
	Link Profilers( 1 )
			
);
2 ACCEPTED SOLUTIONS

Accepted Solutions
ian_jmp

Staff

Joined:

Jun 23, 2011

Solution

Re: jsl colorize profiler and show data in Marginal Model Plots

I'm sure there is a direct way to do this, but I can't recall the syntax at the moment. So here's an indirect way (the 'Wait' statements are for cosmetic effect, of course):

Names Default To Here( 1 );
// Open Tire Tread
dt = Open( "$SAMPLE_DATA/Tiretread.jmp" );
// Use Profiler with predictions
pr = dt << Profiler(Y( :Pred Formula ABRASION, :Pred Formula MODULUS, :Pred Formula ELONG, :Pred Formula HARDNESS ),);
// Assess variable importance
Wait(2);
pr << Dependent Resampled Inputs( 1 );
// Get the 'little red triangle' for Variable Importance
Wait(2);
lrt = Report(pr)[OutlineBox(3)] << getScriptableObject;
lrt << Colorize Profiler;
// Get the 'little red triangle' for Marginal Plots
Wait(2);
lrt = Report(pr)[OutlineBox(10)] << getScriptableObject;
lrt << Data;
mkennke

Community Trekker

Joined:

Mar 18, 2016

Solution

Re: jsl colorize profiler and show data in Marginal Model Plots

Thank you! It works. But my Report is a little bit danamic (one to six Y's). Because of that i modified your script with a "For" and a "Try" function.

Names Default To Here( 1 );
// Open Tire Tread
dt = Open( "$SAMPLE_DATA/Tiretread.jmp" );
// Use Profiler with predictions
pr = dt << Profiler(Y( :Pred Formula ABRASION, :Pred Formula MODULUS, :Pred Formula HARDNESS ),);

// Assess variable importance
pr << Dependent Resampled Inputs( 1 );

// Get the 'little red triangle' for Variable Importance
lrt = Report(pr)[OutlineBox(3)] << getScriptableObject;
lrt << Colorize Profiler;

// Get the 'little red triangle' for Marginal Plots
For (i = 1, i <= 20, i++,
Try(lrt = Report(pr)[OutlineBox(i)] << getScriptableObject;
lrt << Data;))

 

2 REPLIES
ian_jmp

Staff

Joined:

Jun 23, 2011

Solution

Re: jsl colorize profiler and show data in Marginal Model Plots

I'm sure there is a direct way to do this, but I can't recall the syntax at the moment. So here's an indirect way (the 'Wait' statements are for cosmetic effect, of course):

Names Default To Here( 1 );
// Open Tire Tread
dt = Open( "$SAMPLE_DATA/Tiretread.jmp" );
// Use Profiler with predictions
pr = dt << Profiler(Y( :Pred Formula ABRASION, :Pred Formula MODULUS, :Pred Formula ELONG, :Pred Formula HARDNESS ),);
// Assess variable importance
Wait(2);
pr << Dependent Resampled Inputs( 1 );
// Get the 'little red triangle' for Variable Importance
Wait(2);
lrt = Report(pr)[OutlineBox(3)] << getScriptableObject;
lrt << Colorize Profiler;
// Get the 'little red triangle' for Marginal Plots
Wait(2);
lrt = Report(pr)[OutlineBox(10)] << getScriptableObject;
lrt << Data;
mkennke

Community Trekker

Joined:

Mar 18, 2016

Solution

Re: jsl colorize profiler and show data in Marginal Model Plots

Thank you! It works. But my Report is a little bit danamic (one to six Y's). Because of that i modified your script with a "For" and a "Try" function.

Names Default To Here( 1 );
// Open Tire Tread
dt = Open( "$SAMPLE_DATA/Tiretread.jmp" );
// Use Profiler with predictions
pr = dt << Profiler(Y( :Pred Formula ABRASION, :Pred Formula MODULUS, :Pred Formula HARDNESS ),);

// Assess variable importance
pr << Dependent Resampled Inputs( 1 );

// Get the 'little red triangle' for Variable Importance
lrt = Report(pr)[OutlineBox(3)] << getScriptableObject;
lrt << Colorize Profiler;

// Get the 'little red triangle' for Marginal Plots
For (i = 1, i <= 20, i++,
Try(lrt = Report(pr)[OutlineBox(i)] << getScriptableObject;
lrt << Data;))