Other color controls

[markcmiller86]

Is your feature request related to a problem?

No. Feature enhancement.

Is your feature request specific to a data set?

No.

Describe the solution you'd like.

There are some miscellaneous color control issues that kinda sorta seem like they might go with color tables but not sure. That is the selection of color(s) to use for

• missing data
• nan values
• out of bounds values
• void material
• In pos_cmfe we have the concept of the missing overlap value and we specify it as a field value but might want to utilize the notion of a missing value for that too.

Are these the same color concepts as annotation colors (background/foreground, axes, text, etc.)? Should they maybe be treated similarly to how we handle annotation colors?

Depending on circumstance, a user may want colors associated with these cases to be easily disambiguated from "normal" colors from the currently active color table or they may want the colors to be the same as certain colors in the currently active color table.

[biagas]

Out-of-bounds coloring is already available in Pseudocolor plot:

It's controlled in avt/Plotter/avtVariableLegend, which also has the option for setting NanColor as well.

[markcmiller86]

Out-of-bounds coloring is already available in Pseudocolor plot:

Yeah, that is what I was thinking about I think when I added out of bounds there.

But, I was also thinking that these choices conceptually go with the color table or color scheme more than they (should) go with a particular plot (or operator...if we have any operators that involve color selection). Or, maybe we have a concept of a color scheme which embodies these color choices...

Color scheme

• Continuous color table
• Discrete color table
• Nan color
• missing value color
• out of bounds color
• cmfe non-overlap color (we currently handle by value)

[biagas]

The way it is implemented for PC plot directly correlates to how it is handled in VTK.

[markcmiller86]

The way it is implemented for PC plot directly correlates to how it is handled in VTK.

@biagas are you saying we should do it this way because VTK does it this way? Do you find anything unsound about observing that color choices for these kinds of things may live outside any particular plot/operator/expresssion?