Pre-compute pathways for fitting

[mgiammar]

• Implement logic for passing pathways to C code
• Document code changes (docstrings + changelog)
• Update online examples to use sim.optimize()/sim.release()
• Ensure serialization of Simulator does not include pathways

Addresses #134

[mgiammar]

Need to have a check case where there are no pathways for the given spin systems and methods

[codecov]

Codecov Report

Merging #242 (9bc8911) into master (798f170) will not change coverage.
The diff coverage is 100.00%.

@@            Coverage Diff            @@
##            master      #242   +/-   ##
=========================================
  Coverage   100.00%   100.00%           
=========================================
  Files          111       111           
  Lines         7132      7173   +41     
=========================================
+ Hits          7132      7173   +41     

Impacted Files	Coverage Δ
src/mrsimulator/method/__init__.py	100.00% <100.00%> (ø)
src/mrsimulator/method/lib/base.py	100.00% <100.00%> (ø)
...rc/mrsimulator/method/tests/lib_tests/test_base.py	100.00% <100.00%> (ø)
src/mrsimulator/simulator/__init__.py	100.00% <100.00%> (ø)
src/mrsimulator/simulator/tests/test_simulator.py	100.00% <100.00%> (ø)
src/mrsimulator/utils/spectral_fitting.py	100.00% <100.00%> (ø)
tests/test_shift.py	100.00% <100.00%> (ø)

[deepanshs]

The list of arguments for the minimizer is growing. We should split it. Maybe move all args for sim.run() to fcn_kws. For example,

minner = Minimizer(sf.LMFIT_min_function, params, fcn_args=(sim, processor, sigma), fcn_kws={"opt": opt, "n_jobs": 1))

You may also need to add sim.run() related arguments to the sf.LMFIT_min_function