473120_en-US

liliconcom

Fit YbyX Non Parametric Quantile Density Contours

In Fit Y by X, the Non Parametric Quantile Density Contours are super useful for visualizing high density data.

In HLS people use software like FlowJo to make simple bivariate plots that have the data points colored by density. JMP gets close enough, plus there is an option to select rows based on a density range.

It would be super convenient if there was a slider for interactively selecting the range, rather than guessing at ranges until the most appropriate one is discovered iteratively. 

 

User InterfaceRe: Fit YbyX Non Parametric Quantile Density Contours - Status changed to: Acknowledged

@Byron_JMP -Thank you for your suggestion! We have captured your request and will take it under consideration.

Re: Fit YbyX Non Parametric Quantile Density Contours - Status changed to: Investigating

Thanks @Byron_JMP .  We will look into this.  

Here is a prototype script that will do this, I'm sure this could be adapted into an add-in.

Names Default To Here( 1 );
dt = Open( "$SAMPLE_DATA/Big Class.jmp" );
obj = dt << Bivariate( Y( :Weight ), X( :Height ) );
robj=report(obj);
obj << Nonpar Density( {"5% Contours"n( 0 )} );
obj << (curve[1] << "5% Contours"n( 1 ));

sliderLowerValue=0;
sliderUpperValue=1;
robj["Quantile Density Contours"]<< Append(
Panel Box( "Select Points by Non-Parametric Density",
  tb1 = Text Box( "Lower Density: " || Char( sliderLowerValue ) ),
  tb2 = Text Box( "Upper Density: " || Char( sliderUpperValue ) ),
  sb = Range Slider Box(
   0,
   1,
   sliderLowerValue,
   sliderUpperValue,
   tb1 << Set Text( "Lower Density: " || Char( sliderLowerValue ) );
   tb2 << Set Text( "Lower Density: " || Char( sliderUpperValue ) );
   obj << (curve[1] << Select Points by Density( sliderLowerValue, sliderUpperValue ));
  )
 )
 );

 

Re: Fit YbyX Non Parametric Quantile Density Contours

//exactly what I had in mind
//works really well with high density data 

Names Default To Here( 1 );
dt = Open("$SAMPLE_DATA/Cytometry.jmp");
obj = dt << Bivariate( Y( :CD3 ), X( :CD8 ) );
robj=report(obj);
obj << Nonpar Density( {"5% Contours"n( 0 )} );
obj << (curve[1] << "5% Contours"n( 1 ));

sliderLowerValue=0;
sliderUpperValue=1;
robj["Quantile Density Contours"]<< Append(
Panel Box( "Select Points by Non-Parametric Density",
  tb1 = Text Box( "Lower Density: " || Char( sliderLowerValue ) ),
  tb2 = Text Box( "Upper Density: " || Char( sliderUpperValue ) ),
  sb = Range Slider Box(
   0,
   1,
   sliderLowerValue,
   sliderUpperValue,
   tb1 << Set Text( "Lower Density: " || Char( sliderLowerValue ) );
   tb2 << Set Text( "Lower Density: " || Char( sliderUpperValue ) );
   obj << (curve[1] << Select Points by Density( sliderLowerValue, sliderUpperValue ));
  )
 )
 );

Re: Fit YbyX Non Parametric Quantile Density Contours - Status changed to: Yes, Stay Tuned!

Thanks for the suggestion @Byron_JMP, and for developing the script @SamGardner! The developer has agreed to work on this in JMP 18.