Adds plotting methods

[BradyPlanden]

This PR completes the following:

• Adds a cost landscape (plot_cost2D) method
• Default time-series plotting method for comparing
  • Dataset
  • Synthetic model with identified parameters
  • Slider functionality updating the fit for each optimisation step
• Optimiser convergence plot method
• Parameter value per iteration plot method

Update:
I've left the slider off this PR as future functionality to add. Additionally, the above plots work across all optimisers, except plot_parameters() and plot_convergance() which don't currently support NLopt and SciPy optimisers.

[codecov]

Codecov Report

Attention: 14 lines in your changes are missing coverage. Please review.

Comparison is base (5225cfc) 90.57% compared to head (1e505ee) 92.65%.

Files	Patch %	Lines
pybop/plotting/plotly_manager.py	75.47%	13 Missing ⚠️
pybop/optimisers/base_optimiser.py	66.66%	1 Missing ⚠️

Additional details and impacted files

@@             Coverage Diff             @@
##           develop     #115      +/-   ##
===========================================
+ Coverage    90.57%   92.65%   +2.07%     
===========================================
  Files           18       30      +12     
  Lines          647     1007     +360     
===========================================
+ Hits           586      933     +347     
- Misses          61       74      +13     

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

[BradyPlanden]

This should now be ready for review. Four methods have been added:

1. pybop.plot_2d(cost, optim=optim, steps=10) which has an optional argument to overlay the optimiser evolution on the contour and # of steps.
2. pybop.plot_convergence(cost) plots the minimum cost value at each iteration of the optimisation.
3. pybop.plot_parameters(cost) create sub-plots for the evolution of the parameter values during the optimisation.
4. pybop.quick_plot(cost) creates a voltage against time figure for the identified parameters and the target data set.

plot_convergence() and quick_plot() both use the added StandardPlot class for figure creation. This class provide a standard figure implementation for single plot figures, this addition provides a base class for future figure expansion/additions.

All the plotting is completed using plotly, with the setup.py updated to have an optional installation of pip install pybop[plots] or pip install pybop[all].

See updated initial comment on functionality with nlopt and scipy.

Here are some sample figures, generated from the spm_adam.py example. First, quick_plot():

plot_convergence():

plot_parameters():

plot_2d() without optimiser evolution:

and with (where the red marker is the starting position):

[NicolaCourtier]

Nice work on these plotting scripts, thanks @BradyPlanden! The plots are clear and useful - a great addition to PyBOP.