Subscribe Bookmark RSS Feed

Hello, I can't seem to find the JSL code for adding a new column to a current data table that predicts values based on a cubic bivariate fit without a centered polynomial (from data in that table). I can only seem to do it for a linear fit. Also, Is it po

rebecca-maceach

Community Trekker

Joined:

Mar 5, 2015

Hello, I can't seem to find the JSL code for adding a new column to a current data table that predicts values based on a cubic bivariate fit without a centered polynomial (from data in that table). I can only seem to do it for a linear fit. Also, Is it possible in JSL to not actually have the graph populate and to only save the predicted values from it? Thanks!

1 ACCEPTED SOLUTION

Accepted Solutions
Solution

Messages to a fitted curve should be part of a a list, and the Invisible option can be used to hide the report.

dt = Open( "$SAMPLE_DATA/Big Class.jmp" );

dt << Bivariate(

    Y( :weight ),

    X( :height ),

    Fit Special( Degree( 3 ), Centered Polynomial( 0 ), {Save Predicteds} ),

    invisible

);

1 REPLY
Solution

Messages to a fitted curve should be part of a a list, and the Invisible option can be used to hide the report.

dt = Open( "$SAMPLE_DATA/Big Class.jmp" );

dt << Bivariate(

    Y( :weight ),

    X( :height ),

    Fit Special( Degree( 3 ), Centered Polynomial( 0 ), {Save Predicteds} ),

    invisible

);