nav[aria-label="Primary Navigation"] { padding: 0; & ul { list-style: none; width: 100%; display: flex; flex-direction: row; justify-content: start; align-items: start; gap: 30px; padding: 0; & li { margin: 0; } & ul li { list-style: none; } } }

Analysis By/Page Ordering

YGerchman

Hi all,

I have a Partition analysis with "by" column. that means that I have a couple of "pages" (=plots).

obj1=Partition(
Y( col1 ),
X( col2 ),
Show Split Prob( 1 ),
Informative Missing( 1 ),
By( col3));<BR /> part1=obj1<<report;

In addition, I would be glad if I could arrange them in an order (based on the second level probability).

so- I built 2nd table with order number of each table. it looks like this:

part1[x] Probability order
1 0.23 2
2 0.21 1
3 0.52 4
4 0.32 3

how could I show the plots in the order I want? (with or without the 2nd table)

note: If I will send to report or journal - the plots no longer will be interactive which is not good for the application at all.

thank you all,

Yoav.

Accepted answers

YGerchman

the bottom-left one. but in my data - i have only 2 levels (0 or 1).
in any case- i've figure it out by getting partition script and append in into new window.

nw=new window();
for(i,i<nrows(table),i++,
a=obj1[row i from order column ]<<get script;
nw<<append(a));

thanks in any case!

All comments

I don't know of a way to do this interactively except to create a dashboard (maybe someone else does?), but you could make a fairly simple script to order them any way you like. Add a local data filter to your platform and performing the analysis for a single level of your By element, and then copy/paste the script into a New Window script of your own:

Names default to here(1);

//open sample data
dt = Open("$Sample_data/big class.jmp");

//New window and V list box will create a window with each analysis below the previous one
win = New Window("Analysis",
	V List Box(
	
		//copy this script from the platform you used after adding a local data filter
		Bivariate(
			Y( :height ),
			X( :weight ),
			Local Data Filter(
				Close Outline( 1 ),
				Add Filter(
					columns( :age ),
					Where( :age == 14 ),
					Display( :age, N Items( 6 ) )
				)
			)
		),
		
		//then add the comma (previous line), and paste it again, changing the data filter
		Bivariate(
			Y( :height ),
			X( :weight ),
			Local Data Filter(
				Close Outline( 1 ),
				Add Filter(
					columns( :age ),
					Where( :age == 13 ),
					Display( :age, N Items( 6 ) )
				)
			)
		),
		Bivariate(
			Y( :height ),
			X( :weight ),
			Local Data Filter(
				Close Outline( 1 ),
				Add Filter(
					columns( :age ),
					Where( :age == 15 ),
					Display( :age, N Items( 6 ) )
				)
			)
		)
	)
);

//Now, if you want, you can hide all of the data filters:
(win << XPath("//OutlineBox[@helpKey='Data Filter']")) << Visibility("Hidden")

I did just submit a wish list request for one way to make this easier:Respect value ordering in 'By' variable

txnelson

Can you further specify what you mean by "(based on the second level probability)"? Below is a simple run

Names Default To Here( 1 );

//open sample data
dt = Open("$Sample_data/semiconductor capability.jmp");
obj1=Partition(
Y( :wafer ),
X( :pnp1 ),
Show Split Prob( 1 ),
Informative Missing( 1 ))

and here is the output from the following JSL after having performed 1 split on the data.

Which of the tables are you referring to when you ask for the "second level probability"?

YGerchman

the bottom-left one. but in my data - i have only 2 levels (0 or 1).
in any case- i've figure it out by getting partition script and append in into new window.

nw=new window();
for(i,i<nrows(table),i++,
a=obj1[row i from order column ]<<get script;
nw<<append(a));

thanks in any case!