topic Re: Bivariate Fit - how to delete fitting lines and results in Discussions

Bivariate Fit - how to delete fitting lines and results

ohkjuns — Fri, 09 Jun 2023 19:45:14 GMT

I would like to my plots simple after fitting each values. I have searched some solutions, but it is not working to my case.

I want to delete the box containing all of the "Fit Each Value" line items. It looks that it can be done with report(biv)[PictureBox(2)]<<delete;. But unfortunately, "Fit each value" is not "PictureBox". Instead, it looks "Segments" from display trees. How could I address this issue so that make plot looks better?

Re: Bivariate Fit - how to delete fitting lines and results

txnelson — Mon, 26 Apr 2021 09:21:46 GMT

The Fit displays are all located within a Border Box(). If the Border Box() is deleted, the displayed Fits will be deleted..

Report( biv )[Border Box( 2 )] << delete;

Or, you could just "hide" the display by changing it's Visibility

Report( biv )[Border Box( 2 )] << visibility("collapse");

Here is a complete example

Names Default To Here( 1 );
dt = Open( "$SAMPLE_DATA/big class.jmp" );
biv = Bivariate(
	Y( :height ),
	X( :weight ),
	Fit Where( :age == 12, Fit Spline( 1000, {Line Color( {92, 106, 234} )} ) ),
	Fit Where( :age == 13, Fit Spline( 1000, {Line Color( {142, 149, 213} )} ) ),
	Fit Where( :age == 14, Fit Spline( 1000, {Line Color( {192, 192, 192} )} ) ),
	Fit Where( :age == 15, Fit Spline( 1000, {Line Color( {212, 132, 132} )} ) ),
	Fit Where( :age == 16, Fit Spline( 1000, {Line Color( {232, 71, 71} )} ) ),
	Fit Where( :age == 17, Fit Spline( 1000, {Line Color( {252, 11, 11} )} ) )
);
Report( biv )[Border Box( 2 )] << delete;
// or
// Report( biv )[Border Box( 2 )] << visibility("collapse");

Re: Bivariate Fit - how to delete fitting lines and results

ohkjuns — Tue, 27 Apr 2021 17:55:20 GMT

Jim,

I have tested but nothing changed. I have included my script. Could you please take a look and give advice?

Thansk,

Jun Sung

--- coding format broken when I paste it so that reading is not strightforward. please check screen capture in above.

biv = Bivariate(
	Y( :DATA ),
	X( :Rad ),
	By( :Lot7 ),
	SendToReport(
		Dispatch(
			{},
			"Bivar Plot",
			FrameBox,
			{Frame Size( 450, 300 ), Row Legend(
				:WID3,
				Color( 1 ),
				Color Theme( "JMP Default" ),
				Marker( 1 ),
				Marker Theme( "Standard" ),
				Continuous Scale( 0 ),
				Reverse Scale( 0 ),
				Excluded Rows( 0 )
			)}
		),

	),
	Group By( :WID3 ),
	Fit Spline( 0.1, Standardized )
);

Re: Bivariate Fit - how to delete fitting lines and results

txnelson — Tue, 27 Apr 2021 18:04:30 GMT

You left off the statement that does the deletion of the fit displays

Report( biv )[Border Box( 2 )] << delete;

Re: Bivariate Fit - how to delete fitting lines and results

ms — Tue, 27 Apr 2021 18:52:29 GMT

You have a By() column in your analysis which means that your variable biv is not a bivariate platform object but a list of two or more bivariate objects. You will need to loop through them to make the changes to all the sub-reports.

Try this:

xp = (biv << xpath( "//PictureBox" ));
For( i = N Items( xp ), i > 0, i--,
	xp[i][2] << visibility( "collapse" )
);

Re: Bivariate Fit - how to delete fitting lines and results

ohkjuns — Wed, 28 Apr 2021 04:43:13 GMT

There is some deletion with looping. But, it actually deletes legend. How could I edit script to delete fitting lines?

Thanks,

Jun Sung

Before looping script:

With looping script:

Re: Bivariate Fit - how to delete fitting lines and results

ms — Wed, 28 Apr 2021 06:44:33 GMT

If you have a row legend, the fitted-lines box would be the third PictureBox of each bivariate sub-report.

This should work:

xp = (biv << xpath( "//PictureBox" ));
For( i = N Items( xp ), i > 0, i--,
	xp[i][3] << visibility( "collapse" )
);

Re: Bivariate Fit - how to delete fitting lines and results

ohkjuns — Thu, 29 Apr 2021 06:27:57 GMT

Wow. It is great~~~ Could I ask one more question?

Fitting lines are deleted. But, I would like to delete fitting curve results. I have tried [4], but it is not working. Could you please give great advice?

Thanks,

Jun Sung