topic Re: Adding histogram borders to Contour Plot in Discussions

Adding histogram borders to Contour Plot

Anonymous — Fri, 05 Jun 2015 20:41:27 GMT

Hello,

I am wondering if there is a way to add histogram borders to the Contour Plot platform or even within GraphBuilder?

It is a very useful feature within Fit Y by X and was wondering if it is possible or if it has been done before.

Given JSL is fairly flexible I thought it might be do-able but I haven't quite scoped it out.

Thank you,

Re: Adding histogram borders to Contour Plot

louv — Tue, 18 Oct 2016 23:26:34 GMT

I used Application Builder and obtained this view combining Graph Builder and Histograms to mimic what I think you were referring to in the Fit Y by X platform.

Re: Adding histogram borders to Contour Plot

Anonymous — Thu, 11 Jun 2015 04:08:52 GMT

Thanks LouV. This is a nice solution.

I guess, I was wondering if there was a way to add panels/frames within a platform without using ApplicationBuilder to combine them using JSL.

Re: Adding histogram borders to Contour Plot

ian_jmp — Thu, 11 Jun 2015 09:02:32 GMT

You should be able to do what you want (but you still require JSL). Here's a toy example:

dt = Open("$SAMPLE_DATA/Car Physical Data.jmp");

// Contour plot object

cp = dt << Contour Plot(X( :Weight, :Horsepower ), Y( :Turning Circle ));

// Contour plot report layer

cpRep = Report(cp);

// See the structure of the display tree to learn how to modify it

// cpRep << ShowTreeStructure;

// Prepand a new display box to an existing element . . .

Wait(3);

cpRep[ListBox(2)] << Prepend(GraphBox());

// Appand a new display box to an existing element . . .

Wait(3);

cpRep[ListBox(2)] << Append(TextBox("Hello World ... ", << FontColor("Green")));

Re: Adding histogram borders to Contour Plot

johnhoughton — Tue, 18 Oct 2016 23:30:59 GMT

Use copy and paste frame contents

Open( "$ENGLISH_SAMPLE_DATA/Big Class.jmp" );

biv = bivariate( y( height ), x( weight ) , Histogram Borders( 1 ));

cont=Graph Builder(

Show Control Panel( 0 ),

Variables( X( :weight ), Y( :height ) ),

Elements(

Contour( X, Y, Legend( 4 ), Number of Levels( 0 ) ),

Points( X, Y, Legend( 5 ) )

)

);

rbiv = biv << report;

rcont=cont<<report;

rcont[framebox(1)]<< copy frame contents;

rbiv[framebox(3)]<<paste frame contents;;

cont<<close window;

You can do this interactively as well. Create the contour plot , then right click >Edit>Copy Frame Contents.

Then create the bivariate plot with histogram borders, right-click on that and Edit>Paste Frame Contents.

Beauty of this is the axes and frame sizes all adjust automatically,

Another approach isn't so sweet. The code below is incomplete because the axis on both plots should be equal to each other. Demonstates the principal though. (The Wait commands I found were needed to avoid crashing with an exception)

Open( "$ENGLISH_SAMPLE_DATA/Big Class.jmp" );

biv = bivariate( y( height ), x( weight ) , Histogram Borders( 1 ));

cont=Graph Builder(

Size( 308, 270 ),

Show Control Panel( 0 ),

Variables( X( :weight ), Y( :height ) ),

Elements(

Contour( X, Y, Legend( 4 ), Number of Levels( 0 ) ),

Points( X, Y, Legend( 5 ) )

)

);

rbiv = biv << report;

rcont=cont<<report;

wait(.1);

rbiv[framebox(3)]<< sib append(rcont[framebox(1)]);

wait(.1);

rbiv[framebox(3)]<<delete;