Improve PIGs

[JanBenisek]

After some discussion with Geert, below his suggestions:

Incidence graph:

• the colors don't really match with our slide template AND I believe the frequency bars should be less visually attractive (eg grey), while the incidence graph should be really present (blue).
• from this graph it is impossible to guess which Y-axis (left vs right) applies to incidence, and which applies to proportion of cases. Usually, I would align the color of the axis to the graph color - i.e. grey for population size on the left and blue for incidence on the right.
• I would personally put incidence on the left instead of the right, as that is the most important information and typically, we are used to look at the left axis first. I would also clean up the graph labels a little so they don't contain .0 (left, right and bottom)...
• make sure we always refer to them as Predictor Insights Graphs!

Attached is how the plot could look like.

[sandervh14]

Reopening this issue for 2 requested bugfixes:

• the y axis scale can be misleading when the variation on target values is small, for example:
  
• long bin labels get shifted on the x-axis, for example:
  

[hendrikdewinter8]

Should we apply fixed Y axis for all PIG graphs within one analysis?

[sborms]

Fixing the Y-axis is too dangerous as there can be a large Y differential for some bins and small for others. I would (similar to what I saw Sander suggesting elsewhere) take the min-max proportional difference, and when it's small (<5%), we let the Y-axis go from [0.90*min, 1.10*max] (be careful with negative values).

I think that would suffice as a fix. It's never going to be visually perfect for all cases.