Custom cost matrices

[jcoupey]

Issue

This PR aims at pushing the separation between travel times and costs one step further by allowing to pass custom cost matrices in input. Fixes #415.

Tasks

• Parse matrices.*.costs matrices in input
• Adjust vroom::Input to hold the various types of matrices
• Update vroom::CostWrapper to introduce various ways to compute/use costs depending on context
• Define proper cost values when generating vroom::Route objects in solutions
• Fix durations/costs separation leftovers
• Update docs/API.md
• Update CHANGELOG.md
• review

[jcoupey]

Fixes #570

[jcoupey]

While testing, I'm seeing situations where this PR currently hits:

vroom/src/algorithms/local_search/local_search.cpp

Line 1388 in f7dfa76

assert(new_cost + best_gain == previous_cost);

This happens when custom costs and durations matrix are different. It means we have a faulty operator gain evaluation. I suspect this is a leftover from the costs/durations separation done a while back.

[jcoupey]

We're getting there but I'm still seeing weird things going on. For example when using a cost matrix that is the duration matrix doubled, then for some instances the solving path is changed at some point and we get a different solution.

AFAICT this is related to a problem in UnassignedExchange gain evaluation. There are probably still a couple values involving durations instead of costs.

[jcoupey]

After much digging, I think the last problems I'm seeing are not at all cost-related. My feeling is that comparing solutions across instances where we scale the costs is only a way to highlight a problem that is already present in the codebase, so I'll open a dedicated ticket for that.