Draft PR: Remove solver_.parameters_

[montythind]

This is just a draft for the PR.

closes #621

[montythind]

I do see some calls where solver.solver_.parameters_ is not being modifiedbut rather the value is getting used. Example:

EXPECT_EQ(solver.solver_.parameters_.absolute_tolerance_[state.variable_map_["Ar"]], 1.0e-12);
std::vector atol = gpu_solver.solver_.parameters_.absolute_tolerance_;

I am unsure if we need to change these as well?

@K20shores @mattldawson

[codecov-commenter]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 93.34%. Comparing base (16ad525) to head (263af07).
Report is 2 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #668      +/-   ##
==========================================
+ Coverage   92.69%   93.34%   +0.65%     
==========================================
  Files          53       53              
  Lines        3585     3606      +21     
==========================================
+ Hits         3323     3366      +43     
+ Misses        262      240      -22     

Flag	Coverage Δ
	93.34% <100.00%> (?)

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

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[K20shores]

Right direction, but we still have some more work to do. Any mechanism-specific thing (like tolerances) need to be moved somewhere else. We should spend some time figuring out where they need to go. Also, the parameters need to be any set of paramters (rosenbrock, backward euler, etc.)

[K20shores]

Suggested change

	SolverResult Solve(double time_step, StatePolicy& state, RosenbrockSolverParameters params)
	{
	solver_.parameters_.h_start_ = params.h_start_;
	return solver_.Solve(time_step, state);
	}
	SolverResult Solve(double time_step, StatePolicy& state, RosenbrockSolverParameters params)
	{
	solver_.parameters_ = params;
	return solver_.Solve(time_step, state);
	}

We actually need all of the paramters to be updated. That was the impetus for moving the tolerances off of the params. Also, this needs to work for any set of parameters, not just rosenbrock