Extending PWL Costs

[bknueven]

The PWL cost curve handling introduced in #227 fail when the lowest power point on the curve is greater than or equal p_max and when the highest power point on the curve is less than or equal to p_min.

To resolve this, we could either:

1. Determine this is a user error, in which case we should throw a proper error.
2. Handle this case automatically. We already extend the PWL cost curve beyond its given bounds, so extending it in this case is perhaps not the worst thing.

[DLWoodruff]

I need a decision on this issue sooner rather than later. It is a simple question, but I can't honestly say I know what is the right answer. BTW: I think he means "when the lowest power point on the cost curve is greater than p_max " But anyway, the question is: should Egret silently extend the piecewise cost curve?