Added several plot methods and fixed a few minor bugs/warnings

[tg359]

Issues addressed by this PR

Closes #261

See commit message.

Test files

See changes to pytest files for examples of use of the newly added methods. Bug fixes mainly regarded fixing methods which raised warnings about deprecation.

Changelog

Additional comments

[Tom-Kingstone]

There are a couple of missing docstrings, and the utci comparison plot has a formatting issue. I have tested everything that has been changed, and these are the only issues I have come across.

[Tom-Kingstone]

Tested everything again (as well as new stuff added since last comment), and formatting of plots is all fixed now, though there are still some issues with unit tests, and a logic issue with the tilt orientation factor plot method.

[Tom-Kingstone]

Unit tests are failing due to removing create_legend() and heatmap_histogram().
(I see that heatmap_histogram was renamed to annual_heatmap_histogram with practicallly the same function, so that should be fine)

need some more thought on changes to create_legend, maybe keep the original method and just have it call create_legend_handles_labels and then create a legend with the output of that?
see:

_____________________________________________________ test_legend _____________________________________________________

    def test_legend():
        """_"""
>       assert isinstance(TEST_CATEGORICAL_UNBOUNDED.create_legend(), Legend)
E       AttributeError: 'Categorical' object has no attribute 'create_legend'

tests\test_categorical.py:129: AttributeError
_______________________________________________ test_heatmap_histogram ________________________________________________

    def test_heatmap_histogram() -> None:
        """_"""
>       assert isinstance(UTCI_DEFAULT_CATEGORIES.heatmap_histogram(TEST_TIMESERIES_DATA), Figure)
E       AttributeError: 'CategoricalComfort' object has no attribute 'heatmap_histogram'. Did you mean: 'annual_heatmap_histogram'?

tests\test_categorical.py:182: AttributeError

[Tom-Kingstone]

why shuffle the iterations?
I would expect the order doesn't matter much, though when not running parallel it might be better if they were in order so that if one fails/is missed, it's easier to tell which one it was.

[Tom-Kingstone]

This path also needs to include the analysis period description, as the analysis period is taken into account when creating the sky matrix.
Or some sort of representation of analysis_period.hoys, though I think ap start and end dates/times would be best.

As it currently is, if using the same tilts, directions and epw file, but different analysis periods (eg. comparing summer and winter), then one analysis period would be pickled, and then future analysis periods would only show the first used, ignoring the rest.

[Tom-Kingstone]

Does this replace the functionality of plot_tilt_orientation_factor? If so, remove that method as this is just an improvement upon that old version.