topic Re: Distribution analysis (e.g. summarize statistics) direct calculation after selection? in Discussions

Distribution analysis (e.g. summarize statistics) direct calculation after selection?

Fabian88 — Tue, 27 Jun 2017 09:43:15 GMT

Dear all,
I´m sure its an easy question - so for you I think its an quick and easy answer :)

I would like to calculate the data in the distribution-overview based on the selection:

Here for example I would like to see the summary statistics direct after selecting this data (new table calculation). For now I just can see this specific calculation if I Hide and Exclude the unneccesary data in the raw-data-table.

Is this possible?

Im looking forward to your answer. Thanks alot!
Fabian

Re: Distribution analysis (e.g. summarize statistics) direct calculation after selection?

ron_horne — Tue, 27 Jun 2017 10:52:20 GMT

perhaps the following,

once you selected the subset go to Tables>>>subset and produce a subset with only the selected rows.

then, if you open the distribution platform and hit the recall button you should get what you are looking for.

Ron

Re: Distribution analysis (e.g. summarize statistics) direct calculation after selection?

Fabian88 — Tue, 27 Jun 2017 11:47:58 GMT

Hey ron,
Thank you for your fast response.
With your solution I have a distribution with just the subset-data. This solution is the same compared to "Hide and Exclude" the unneccesary data.
What I am looking for is the same distribution overview, and if i select a subset (e.g. M/F) i want the calculated data based just on this subset.

Re: Distribution analysis (e.g. summarize statistics) direct calculation after selection?

ian_jmp — Tue, 27 Jun 2017 12:10:48 GMT

Perhaps you can use the local data filter?

NamesDefaultToHere(1);
dt = Open("$SAMPLE_DATA/Big Class.jmp");
dist = dt << Distribution(
					Automatic Recalc( 1 ),
					Continuous Distribution( Column( :height ) ),
					Continuous Distribution( Column( :weight ) ),
					Local Data Filter(
						Add Filter(
							columns( :age, :sex ),
							Display( :age, Size( 160, 90 ), List Display )
						)
					)
				);

Alternatively, you could use a dashboard with a selection filter

Re: Distribution analysis (e.g. summarize statistics) direct calculation after selection?

txnelson — Tue, 27 Jun 2017 12:40:03 GMT

Or are you looking to click on the Male/Female distribution, and for it to change the values in the other graphs

Here is an example from the Scripting Index using a Data Filter Source Box

Names Default To Here( 1 );
dt = Open( "$SAMPLE_DATA/Big Class.jmp" );
New Window( "Selection Filter",
	Data Filter Context Box(
		H List Box(
			Data Filter Source Box(
				Graph Builder(
					Size( 208, 207 ),
					Show Control Panel( 0 ),
					Show Legend( 0 ),
					Variables( X( :age ) ),
					Elements( Bar( X, Legend( 3 ) ) ),
					SendToReport(
						Dispatch(
							{},
							"Graph Builder",
							OutlineBox,
							{Set Title( "Filter" )}
						)
					)
				)
			),
			Platform(
				Current Data Table(),
				Bubble Plot(
					X( :weight ),
					Y( :height ),
					Sizes( :age ),
					Title Position( 0, 0 )
				)
			)
		)
	)
);

Here is the code adapted to the example you gave

Names Default To Here( 1 );
dt = Open( "$SAMPLE_DATA/Big Class.jmp" );
New Window( "Selection Filter",
	Data Filter Context Box(
		H List Box(
			Data Filter Source Box( Distribution( Nominal Distribution( Column( :sex ) ) ) ),
			Platform(
				Current Data Table(),
				Distribution(
					Nominal Distribution( Column( :age ) ),
					Continuous Distribution( Column( :height ) ),
					Continuous Distribution( Column( :weight ) )
				)
			)
		)
	)
);

Re: Distribution analysis (e.g. summarize statistics) direct calculation after selection?

Fabian88 — Tue, 27 Jun 2017 12:52:52 GMT

@ian_jmp - thx. the local data filter is a good way to select the data!
@txnelson - I like your solution over the Data Filter Source Box as well.. i rebuild your solution by using your script! But can you show me the solution how i can set up this directly over the jmp software? This would be handy to show this to my College.

Re: Distribution analysis (e.g. summarize statistics) direct calculation after selection?

txnelson — Tue, 27 Jun 2017 12:57:04 GMT

I am not understanding what you are asking.

" But can you show me the solution how i can set up this directly over the jmp software?"

Could you provide more details?

Re: Distribution analysis (e.g. summarize statistics) direct calculation after selection?

Fabian88 — Tue, 27 Jun 2017 13:03:16 GMT

Sry..
If i run your script based on my data, I get the "Data Filter Context Box". If i select M I get the new graphs and as well the new calculated "Summary statistics".
Thats excatly how i want it.
My last question here is: How can i set up this without using the script? Is there a solution which i can select in the jmp-software? I want to explain my collegue to set this up without using the script (we don´t have programming skills ;-))
Do you understand what I want?

Re: Distribution analysis (e.g. summarize statistics) direct calculation after selection?

txnelson — Tue, 27 Jun 2017 13:30:28 GMT

The Dashboard facility in JMP has a template that would allow you to setup such filtering..

New==>Dashboard==>Hierarchical Filter

Re: Distribution analysis (e.g. summarize statistics) direct calculation after selection?

ian_jmp — Tue, 27 Jun 2017 13:45:49 GMT

One of the great things about Dashboard Builder is that you don't need any programming. JMP add-ins also make it easy to disseminate dashboards to others.

Just as an example, I made the add-in attached in 30 seconds, with no coding. If you double click on it to install, you can then go to the 'Add-Ins' menu and select 'Dashboard' to run it. It will look for file 'Big Class.jmp', which you can find in the sample data.