Non-communicative data filter after sorting rows

[#1]

I have a window with a dynamic data filter. The filter is used to distill a table down before going to an external python function for post processing. The filter dynamically updates well when deleting rows. However, if I sort the rows in my data table by certain column values it seems like the table is disconnected /non-communicative with the window since changing the filter no longer triggers a python call. Even if I undo the sorting it still persists. I have a feeling that this is due to the fact that sorting function replaces the current data table and it may be necessary to somehow resync the filter.

dt = Current Data Table();

// Python execution function
run_python_func = Function( { dt },
	Python Send( dt );
	Python Execute(
		{ },
		{ output },
		"\[
import jmp

# Do some stuff
output = 4 + 4
		]\"
	);
   Show( output );
);

plot_window = New Window( "Python post process",
	Data Filter Context Box(
		H List Box(
			cb = Check Box( "Enable real-time filtering" ),
			filter = (dt << Data Filter( Local, Add Filter( columns( :DIE_ID, :MACRO, :TEST_NAME ) ) ) )
			)
		)
	)
);

// Do real-time filtering
refresh_filter = Function( { do_filt },
	If( do_filt,
		// Get selected rows and filter input data table
		rows = filter << get filtered rows;
		dt_filt = dt << Subset( Rows(rows), selected columns(0) );
		dt_filt << Show Window(0);
		
		// Run python function
		run_python_func( dt_filt );
		Close( dt_filt, NoSave);
	);
);

rs = filter << Make Filter Change Handler( refresh_filter( cb << Get() ) );

[#2]

Data filter will get "disconnected" from the table. If you just create a global data filter to a table and then try to sort it, JMP will give you following message

Names Default To Here(1); 

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

plot_window = New Window("Python post process",
	Data Filter Context Box(
		H List Box(
			cb = Check Box("Enable real-time filtering"), 
			filter = (dt << Data Filter(Global, Add Filter(columns(:age, :sex, :name))))
		)
	)
);

// Do real-time filtering
refresh_filter = Function({do_filt},
	show(Today());
);

rs = filter << Make Filter Change Handler(refresh_filter(cb << Get()));

Write();

There might be some ways around this, but implementation will depend on your use case.

[#3]

Deleting rows works fine, but many other changes to the data table aren’t allowed when reports are open or when a linked subset exists. To be honest, this is quite unexpected for JMP "19" — it feels like 199x.

JMP is already powerful—but being able to freely modify a data table even when reports or linked subsets exist would be a huge productivity boost.

In particular:

• sorting the data table
• adding rows
• concatenating tables

should be possible without forcing the user to manually tear down the analysis context:Sort data table + add rows + concatenate -- make it possible 

If you’re running into the same problem, just follow the link and support the wish with a Kudo *)

At the very least, JMP should:

• Make dependencies transparent
  • What exactly is blocking the operation?
  • Which report, subset, or data filter holds the lock?
• Provide a clear way to apply the requested change
  • e.g., by offering to close and re-open the affected report(s) or subset(s) automatically

*) NB: Unfortunately, the Kudos here don't affect the status of the wish.