Share your ideas for the JMP Scripting Unsession at Discovery Summit by September 17th. We hope to see you there!
Choose Language Hide Translation Bar
Highlighted
andrewtkarl
Level III

Cannot ignore preferences in Fit Model?

I am having trouble using Ignore Platform Preferences() with Fit Model. In this example, I turn on AICc in the preferences, and then try to use Ignore Platform Preferences to run Fit Model without the AICc output, but it still appears. Is there something wrong with my syntax?

 

As a bonus question, is there a more global option to ignore all preferences and platform preferences when running a script? Almost in the same way that names default to here(1) creates a unique namespace for that script, an option to ignore all user preferences when running platforms in a script?

 

Thanks.

 

set platform preferences(Fit Least Squares(AICc(1)));
get platform preferences(Fit Least Squares);
open("$sample_data/big class.jmp");
Fit model(
	Ignore Platform Preferences( 1 ),
	Y( :height ),
	Effects( :weight ),
	Personality( "Standard Least Squares" ),
	Emphasis( "Effect Leverage" ),
	Run()
);
1 REPLY 1
Highlighted
gzmorgan0
Super User

Re: Cannot ignore preferences in Fit Model?

@andrewtkarl,

It seems there are several analysis options that have default settings.  For example, if the following script is run for data table Big Class. jmp, the result will display only those preferences that I have set and saved.  Note I do not have Confidence Diamonds checked. However, if from the UI I select the option Means/Anova/Pooled t, the result will include Confidence Diamonds,the internal defaults for that analyses.

 

I believe for the platform Fit Model, Personality( " Standard Least Squares") acts similarly, however Fit Least Squares preferences has more control.  I suggest scripters use Ignore Platform Preferences(1) and make sure they turn on the ones they want and off the ones that are internal defaults for an analyses (such as, Confidence Diamonds or AICc).

 

Oneway(
	Y( :weight ),
	X( :sex )
)

Even interactively, create a custom least squares Fit model that does not specify the model factors, just the desired report options and add it to a custom menu.  Then you get the usual launch window, however the analyses will have what you want without having to use the menu options to customize the report.

 

Try this script with Big Class.jmp or another data table opened, and you will see how it behaves. 

 

 

Fit Model(
	Ignore Platform Preferences( 1 ),
//	Y( :height ),
//	Effects( :weight ),
	Personality( "Standard Least Squares" ),
	Emphasis( "Effect Leverage" ),
	Summary of Fit( 1 ), Analysis of Variance( 1 ),
		Parameter Estimates( 1 ), Scaled Estimates( 0 ),
		Plot Actual by Predicted( 1 ), Plot Residual by Predicted( 1 ),
		Plot Studentized Residuals( 0 ), Plot Effect Leverage( 1 ),
		Plot Residual by Normal Quantiles( 0 ), Box Cox Y Transformation( 0 ),
		AICc( 0 )
			
);

 

I do not know of a global ignore preferences command.  However, note if you take your original script and disable AICc, AICc(0), for Fit Least Squares, when you run your script, AICc is not reported.

 

set platform preferences(Fit Least Squares(AICc(0)));
get platform preferences(Fit Least Squares);
open("$sample_data/big class.jmp");
Fit model(
	Ignore Platform Preferences( 1 ),
	Y( :height ),
	Effects( :weight ),
	Personality( "Standard Least Squares" ),
	Emphasis( "Effect Leverage" ),
	Run
);

Article Labels

    There are no labels assigned to this post.