Allow generic solver to be passed to Newton method

[rsenne]

This PR is in reference to issue #1170. To solve this I have edited the interface to allow a user to pass a generic solver. This in theory would allow one to use any matrix type given the user has that package loaded and there exists a solver designed for that matrix type on board. This also allows one to write a custom solver (i.e., problem specific factorization, strategies for solving) to circumvent the default behavior. I have included some examples of this in testing including:

1. Different factorizations (would allow users to circumvent the default PositiveFactorizations.jl implementation of cholesky)
2. Usage of a non-originally supported matrix type (StaticArray)
3. Test type preservation of Hessian with TwiceDifferentiable object using BlockArrays.jl
4. A more realistic example of a non-trivial use case (block-tridiagonal hessian) this implementation is not totally optimized as it is a proof of principle

Let me know if there are any other tests you would like me to include or changes to the documentation.

[github-actions]

Benchmark Results

	master	330ffb4...	master / 330ffb4...
multivariate/solvers/first_order/AdaMax	0.652 ± 0.0083 ms	0.647 ± 0.008 ms	1.01 ± 0.018
multivariate/solvers/first_order/Adam	0.652 ± 0.008 ms	0.647 ± 0.008 ms	1.01 ± 0.018
multivariate/solvers/first_order/BFGS	0.224 ± 0.0037 ms	0.234 ± 0.0069 ms	0.955 ± 0.032
multivariate/solvers/first_order/ConjugateGradient	0.0481 ± 0.00054 ms	0.0502 ± 0.00064 ms	0.958 ± 0.016
multivariate/solvers/first_order/GradientDescent	1.71 ± 0.016 ms	1.79 ± 0.007 ms	0.956 ± 0.0098
multivariate/solvers/first_order/LBFGS	0.219 ± 0.0037 ms	0.228 ± 0.0051 ms	0.957 ± 0.027
multivariate/solvers/first_order/MomentumGradientDescent	2.51 ± 0.018 ms	2.59 ± 0.011 ms	0.968 ± 0.008
multivariate/solvers/first_order/NGMRES	0.553 ± 0.011 ms	0.554 ± 0.011 ms	0.999 ± 0.028
time_to_load	0.529 ± 0.0013 s	0.53 ± 0.0063 s	0.999 ± 0.012

Benchmark Plots

A plot of the benchmark results have been uploaded as an artifact to the workflow run for this PR.
Go to "Actions"->"Benchmark a pull request"->[the most recent run]->"Artifacts" (at the bottom).

[pkofod]

Thanks, I'll review :)

[codecov]

Codecov Report

Attention: Patch coverage is 90.00000% with 1 line in your changes missing coverage. Please review.

Project coverage is 85.69%. Comparing base (6850998) to head (330ffb4).

Files with missing lines	Patch %	Lines
src/multivariate/solvers/second_order/newton.jl	90.00%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #1176      +/-   ##
==========================================
- Coverage   85.70%   85.69%   -0.02%     
==========================================
  Files          46       46              
  Lines        3596     3593       -3     
==========================================
- Hits         3082     3079       -3     
  Misses        514      514              

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

🚀 New features to boost your workflow:

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