## How can modify the below 3 factors of 3 levels RSM script to 3 factors of 4 levels?

Community Trekker

Joined:

Oct 21, 2015

DOE(

Response Surface Design,

{Add Response( Maximize, "Y", ., ., . ), Change Factor Settings( 1, -1, 1, "X1" ),

Change Factor Settings( 2, -1, 1, "X2" ),

Add Factor( Continuous, -1, 1, "X3", 0 ), Set Random Seed( 657142714 ),

Make Design( 2 ), Set Axial Choice( 3 ), Center Points( 2 ), Simulate Responses,

Make Table}

);

4 REPLIES

Super User

Joined:

Jul 13, 2011

Often the start point of making a script is to do it interactively in JMP.  So you might want to try creating the design in JMP interactively first them capturing the script.  But here is the snag - your script is creating a central composite design.  These inherently have 5 levels unless the axial points become face-centred in which case you have 3 levels.  I'm not aware of a classical RSM design with 4 levels.  If the need for 4 levels is due to having a categorical factor then you might want to look at using a custom design.

-Dave

Staff

Joined:

Jun 23, 2011

Building on what David has mentioned here is a 16 run custom design driving the 4 levels by utilizing the discrete numeric capability.

DOE(

Custom Design,

{Add Response( Maximize, "Y", ., ., . ),

Add Factor( Discrete Numeric, {1, 2, 3, 4}, "X1", 0 ),

Add Factor( Discrete Numeric, {1, 2, 3, 4}, "X2", 0 ),

Add Factor( Discrete Numeric, {1, 2, 3, 4}, "X3", 0 ),

Set Random Seed( 1212745 ), Number of Starts( 9260 ), Add Term( {1, 0} ),

Add Term( {1, 1}, {2, 1} ), Add Term( {1, 1}, {3, 1} ),

Add Term( {2, 1}, {3, 1} ), Set Sample Size( 16 ), Optimality Criterion( 2 ),

Make Design}

)

Community Trekker

Joined:

Oct 21, 2015

Thanks David

Thanks Lou,

There is a question might be asked that is on what basis does custom design calculate the number of runs?

Ahmed,

Staff

Joined:

Jun 23, 2011

Ahmed,

The answer to your question can be found searching on the internet (JMP custom design number of runs).

http://www.jmp.com/support/notes/48/962.html

In general the number of runs is driven by the model that one specifies. In your example I chose custom design with 3 X 4 level discrete factors and specified all main effects, two-way interactions and polynomial terms.

Also custom design allows one to specify any number of runs between the minimal and recommended number or even exceeding the recommended. In my experience the number of experiments is often the driver due to expense or material availability.