Visualize models with single-peaked criteria

[jacquev6]

As of lincs 2.0, lincs visualize classification-model does not work with models with at least one single-peaked criterion.

The main reason for that is my (@jacquev6) lack of enthusiasm for the ideas we have so far. I'll try an detail them in comments below. The purpose of this discussion is to gather ideas from anyone who might have some hopefully better ones.

[jacquev6]

Idea by Vincent Mousseau:

Instead of coloring a surface for each category, we color intervals on each criterion axes.

Pros:

• probably relatively easy to implement using Matplotlib

Cons:

• we loose the visual impact of the category surfaces

[jacquev6]

In that case, we could even use bars like @manonCABA suggested:

[jacquev6]

We keep surfaces for categories, and bend the axes of single-peaked criteria.

Pros:

• we preserve the visual impact of the category surfaces

Cons:

• impossible (?) to implement using Matplotlib
• the bent axes might be difficult to understand