325474_en-US

liliconcom

Adding calculated columns to the Process Screening summary table with JSL

I swear I saw a post somewhere where someone was directly adding calculated columns to the process screening summary, and I was wanting to do that.

 

The specific column I wanted to add would be just a difference between the Specified Sigma and the Within Sigma, although if I can add one it won't be long before I want to add quite a few more...

 

Anyways, anyone know how to programmatically add additional calculated columns, or if that is possible?

AnalyticsScriptingRe: Adding calculated columns to the Process Screening summary table with JSL

The more I play with this tool the more I am feeling like this is one of the most important additions to JMP since graph builder.  There are so many things I want to continue to add to it.   Like changing my marker type in the goal plot based on what property (or whatever grouping level).  I think this would also be doable with some scripting, may look at that as well.

 

It's funny because (years ago) I used to beat up on JMP just a bit for being a little behind on the quality side, but it has made incredible leaps and bounds, and I would argue that with this platform is now an industry leader.

Re: Adding calculated columns to the Process Screening summary table with JSL

I am glad you asked because I didn't include the fix for column ordering in my last post.  I updated the script in that reply to write a column to the table and then delete it before pulling the values to be calculated. This fixes the sort order problem, but I hope JMP can implement a better solution in the next version of JMP.

Re: Adding calculated columns to the Process Screening summary table with JSL

This was the one I was looking for.  Quick question though, are you saying that any results out of this won't sort?  That will be a bit of a show stopper for me on this one.

Re: Adding calculated columns to the Process Screening summary table with JSL

This might be the post you referred to.  I am still looking for a better way make sure the table sorts correctly when adding the new column, but it does the job: Process Variable Drift or Creep Detection 

names default to here(1);
dt = Open("$Sample_Data/iris.jmp");

//Open Process Screening with drift summaries
ps = dt << Process Screening(
 Y( :Sepal length, :Sepal width, :Petal length, :Petal width ),
 Control Chart Type( "Indiv and MR" ),
 Drift Summaries( 1 )
);

//Add a dummy column and use set values to reset the table order
ncb = (ps << XPath( "//TableBox" )) << Append( newCol = NumberColBox("Delete Me", {}));
newCol << Set Values({});
newCol << Delete Box;

//calculate the difference in sigmas
WithinSigma = (ps << XPath( "//NumberColBox[NumberColBoxHeader[text()='Within Sigma']]" ))[1] << Get;
SpecifiedSigma = (ps << XPath( "//NumberColBox[NumberColBoxHeader[text()='Overall Sigma']]" ))[1] << Get;
SigmaDifference = Matrix(WithinSigma) - Matrix(SpecifiedSigma);

//Add column to table with difference in sigma
ncb = (ps << XPath( "//TableBox" )) << Append( NumberColBox("Sigma Difference", SigmaDifference) );