Frvcpy

This release addresses a handful of minor concerns raised during the peer review process.

---

The primary changes in this patch include

1. The addition of a command line argument to verify that the triangle inequality holds for a passed instance. This is currently a required condition for the algorithm's proper execution, although this may change in the future.
2. Improved testing. Some users running Python 3.7 reported floating point precision issues when running `frvcpy`'s test suite. These issues have been resolved.

frvcpy: first (pre-)release

frvcpy determines the optimal insertion of charging operations into an electric vehicle's route, a problem known as the FRVCP (fixed route vehicle charging problem).

The FRVCP often arises in routing problems for EVs, since its solution is necessary to determine the actual duration of a candidate route.

frvcpy uses the labeling algorithm from [Froger et al. (2019)](https://www.sciencedirect.com/science/article/abs/pii/S0305054818303253), offering optimal solutions in low runtime.

With this initial release, frvcpy has native support for:
- realistic charging functions
- different types of charging stations (CSs)
- compatibility with the [VRP-REP](http://www.vrp-rep.org/) format
- inserting multiple CSs between stops in the route
- route duration constraints
- customer processing times

Future releases may include support for:
- discrete charging decisions
- CS waiting times
- customer time windows
- multi-graph support
- user-specified numerical precision
- more flexible input, more verbose output
- and more!

Install via `pip install frvcpy` and run either in Python or the command line.

Feature requests are encouraged, as are contributions.

Happy EV routing.