Gradient Settings with Column Switcher

[hogi]

☐ cool new feature
☑ could help many users!
☑ removes something that feels like a „bug“
☐ nice to have
☐ nobody needs it

What inspired this wish list request?
When using Column Switcher in Graph Builder, it is common to switch between many measured variables that have completely different value ranges. A color scale that is meaningful for one column is often useless for another column. Currently, color scale settings appear to be shared across switched columns, which frequently results in misleading visualizations and requires manual adjustment after every switch.

What is the improvement you would like to see?
Please support meaningful per-column color scaling when using Column Switcher.

Preferred solution:

1. Column Properties for Color Scales
   • Allow columns to store preferred color scale settings, similar to existing axis-related properties.
2. Robust Automatic Scaling
   • Add a robust range option based on:
     • Median ± N × MAD
   • This would provide useful color scales without being dominated by a few extreme outliers.
     
     
3. Column-Specific Column Switcher Settings
   • Allow Column Switcher to remember color scale settings separately for each switched column.
   • For example, each column could retain its own Min, Max, Range Type, and legend settings.

Why is this idea important?
Column Switcher is intended to simplify exploration of many different variables. However, users currently spend unnecessary time correcting color scales whenever variables have different distributions.

Supporting column-specific color scaling would:

• Reduce manual report maintenance.
• Improve interpretation of heat maps and wafer maps.
• Prevent misleading visualizations caused by inappropriate ranges.
• Make reports more reusable across different data sets.
• Create behavior that users naturally expect when switching between columns with different measurement scales.
• make JMP feel like JMP already feels at many other places: smooth, helpful, no workarounds needed.

The result would be more meaningful visualizations, better workflows, and fewer situations where the current behavior feels like a bug.

other wishes

[hogi]

example:

adjust the gradient for NPN1 while keeping meaningful plots for the other parameters

Open( "$SAMPLE_DATA/Semiconductor Capability.jmp" );
:NPN1[1370::1455]= 5000;
Graph Builder(
      Variables( X( :lot_id ), Y( :wafer ), Color( :NPN1 ) ),
      Elements( Points( X, Y, Legend( 7 ) ) ),
      Column Switcher( :NPN1, {:NPN1, :SPM1, :PNP7, :IVP4, :PBA1, :SIT2} ),
      SendToReport(
            Dispatch( {}, "400", ScaleBox,
                  {Legend Model(
                        7,
                        Properties(
                              0,
                              {gradient(
                                    {/*Scale Values( [110 120] ), */Range Type( "Middle 90%" ),
                                    Label Format( "Fixed Dec", 12, 1 )}
                              )},
                              Item ID( "NPN1", 1 )
                        )
                  )}
            )
      )
);