Corner plotting gives unhelpful error if refinement is run with too few steps

EDIT: the solution to this problem is by replacing the `corner` code with skopt's own plotting:

**jfkcooper (from comment below):**
>I found out (after writing all these methods) that there are actually some nicer ways of plotting the outputs already built into sk-opt. I would suggest using one/multiple of these and replacing some of my plotting code.
https://scikit-optimize.github.io/stable/auto_examples/plots/visualizing-results.html
using `from skopt.plots import` [`plot_objective`](https://scikit-optimize.github.io/stable/modules/generated/skopt.plots.plot_objective.html#skopt.plots.plot_objective)
---
If a refinement is run with fewer than 10 steps, the following error message appears upon running `control.plot_results()`:

```
Your data file apears not to have any points in, please check you have run the refinement and it saved correctly.

---------------------------------------------------------------------------
TypeError Traceback (most recent call last)
Cell In[9], line 3
1 # Run the refinement, i.e. refine the FF parameters against the data
2 control.refine(n_steps=8)
----> 3 control.plot_results();

File /usr/local/lib/python3.11/site-packages/MDMC/control/control.py:556, in Control.plot_results(self, filename, points, MH_norm)
553 filename = self.results_filename
554 plotter = PlotResults(filename, MH_norm=MH_norm, points=points,
555 quantiles=[0.34, 0.5, 0.68])
--> 556 cornerplot, means, stds = plotter.create_cornerplot()
558 if self.verbose != -1:
559 print(f'Parameter means = {means}, Parameter errors = {stds}')

TypeError: cannot unpack non-iterable NoneType object
```
**To reproduce:** this can be done with the Argon a-to-z tutorial; try it with fewer than 10 steps, then greater than 10 steps.

What is actually happening is if the method `_expected_minimum_random_sampling` produces an `IndexError`, the `create_cornerplot` function returns None and this makes `Control.plot_results` fall over. This *can* happen if the results file is not found, but can also happen if it isn't long enough!

The error handling should be improved here:
- if the corner plot actually fails to generate (i.e. there's too few datapoints to create valid contours), then this should be the error message.
- if the data file *doesn't* have any points in, then this should be an error itself: currently the exception handling is
```python
except IndexError:
msg = ("\n \n Your data file apears not to have any points in, please check you have "
"run the refinement and it saved correctly. \n")
print(msg)

return None
```
and then the None trips up a python error elsewhere, when it really should be
```python
except IndexError as error:
raise IndexError("Failed to read any points from the results file. Please check you have "
"run the refinement and it saved correctly.") from error
```
unless there's a really good case where this should be returning None to a user expecting a 3-tuple, rather than crashing out immediately.

- Alternatively, control.plot_results() could be rewritten to use the minimizer history directly when plotting results, rather than the results file, which would sidestep this exception needing to exist at all.