implement evaluation of piecewise curves

[luke-kiernan]

Evaluation as a function, for PiecewiseIncrementalCurve and PiecewiseAverageCurve. I also added check_domain(fd::FunctionData, x::Real), as defense against floating point errors.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 80.41%. Comparing base (0ee9770) to head (d67d939).

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #514      +/-   ##
==========================================
+ Coverage   79.10%   80.41%   +1.30%     
==========================================
  Files          71       71              
  Lines        6192     6141      -51     
==========================================
+ Hits         4898     4938      +40     
+ Misses       1294     1203      -91     

Flag	Coverage Δ
unittests	80.41% <100.00%> (+1.30%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

Files with missing lines	Coverage Δ
src/cost_aliases.jl	92.00% <100.00%> (+6.63%)	⬆️
src/function_data.jl	99.57% <100.00%> (+1.30%)	⬆️

... and 29 files with indirect coverage changes

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[GabrielKS]

The current approach duplicates already-implemented FunctionData and ValueCurve conversion math. @luke-kiernan and I are in the midst of a conversation about what we even want the interface to look like, I'm going to mark this as draft until we resolve that.

[luke-kiernan]

We've agreed that it makes sense to evaluate an InputOutputCurve at an input x: just evaluate the function data at x. However, it's ambiguous for the Incremental and AverageRate curves, so there, I'll leave evaluation un-implemented. (If you want to evaluate the curve--not the rate--for one of those, just convert it to an InputOutputCurve first.)

Will do in a separate PR.

[GabrielKS]

Successor: #517