cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 
Try the Materials Informatics Toolkit, which is designed to easily handle SMILES data. This and other helpful add-ins are available in the JMP® Marketplace
Choose Language Hide Translation Bar
lisa2442
Level I

How do I script DOE Evaluate Design over a variable array of X Factors?

I have a script where after opening a data table, the user gets to select which columns are the X regressors/factors:

Panel Box( "Column Selection",
    Lineup Box( N Col( 2 ),
        Text Box( "Select Controlled Conditions" ),
        cont_cond list = Col List Box( All )
    )
)

 

The list cont_cond has worked very well for me in the rest of my script (creating Exploratory Data Analysis and saving the output graphs/distributions/etc.. to a Journal). 

 

However, now I am on my last task and am completely stumped. The GOAL would be to create a Color Map of Correlations for all main effects and two-way interactions for all of the elements in the cont_cond array. I am not creating a new DOE. This is for evaluation of data that already exists in a data table. This is as far as I've got:

 

d = DOE( Evaluate Design, X( cont_cond( 1 ) ) );

 

This at least brings up the Evaluate Design box and (not sure why since I only selected the (1) element), loads all the elements from my cont_cond array into the "Factors" box. I can manually select the Main Effects and Interactions in the Model box and then manually send the resultant Color Map on Correlations to the Journal that has already been opened with my previous script; but the idea is for it to be automatically added. 

 

Again, all the research I have done on this community page and JSL Scripting Index only have examples for hard coding in each named or indexed X of a fixed list of factors for the argument above. I also abandoned the Analyze - Multivariate platform as it does not let me bring my categorical X factors in. 

 

1. How do I write the d = DOE( Evaluate Design, X???) code so that all of the elements in the cont_cond array are selected for the Design Evaluation?

2. How do I also ensure the two-factor interactions (which will also be variable based on the number of elements in the cond_cont array) are also selected? 

3. Once this is accomplished, how do I then script the journaling of the resultant Color Map on Correlations vs. manually right clicking to Journal? 

 

Many thanks!

Lisa

 

1 REPLY 1
tbidwell
Level III

Re: How do I script DOE Evaluate Design over a variable array of X Factors?

I don't know if this helps any, but you should be able to grab the Correlations Map from the Evaluate DOE report using something like what I've shown below.  I am using some data from the sample data library to illustrate.

 

names default to here(1);
dt = open("$sample_data\quality control\thickness.jmp");

eval = DOE(Evaluate Design, x(:thickness 01, :thickness 02, :thickness 03));
reval = eval << report;
reval << show window(0); //here to minimize the Evaluate DOE report window
colormap = reval["Color Map on Correlations"];  //this grabs the Color Map Outline box

new window("test",
	Outline box("correlations map", colormap)
);