BUG: Bug #10386 - improved handling of inf/nan.
The main issue with inf/nan seems to be in calculating the bounds of the points. I have created a structure to store a list of bad points, and then a function to calculate the bounds ignoring these bad points. After talking to Utkarsh we decided that bad points should be kept, and breaks in the line plot should appear. This will require further testing, and possibly some extra logic in the rendering code (works for me).
Showing with 144 additions and 4 deletions