remove _ensure_plottable

[dschwoerer]

The plotting backend does more reliable checking and thus removing avoids
false negatives, which are causing easily avoidable plot failures

• Closes Plotting of labelled data fails #5762
• Tests added
• Passes pre-commit run --all-files
• User visible changes (including notable bug fixes) are documented in whats-new.rst
• New functions/methods are listed in api.rst

[pep8speaks]

Hello @dschwoerer! Thanks for updating this PR. We checked the lines you've touched for PEP 8 issues, and found:

• In the file xarray/plot/utils.py:

Line 4:1: F401 'datetime.datetime' imported but unused

Comment last updated at 2021-10-24 10:33:03 UTC

[github-actions]

Unit Test Results

         6 files           6 suites   53m 30s ⏱️
16 209 tests 14 462 ✔️ 1 736 💤 11 ❌
90 450 runs  82 204 ✔️ 8 180 💤 66 ❌

For more details on these failures, see this check.

Results for commit 8ecbb39.

♻️ This comment has been updated with latest results.

[dcherian]

Thanks @dschwoerer This makes sense to me.

A lot of failed tests are checking for the raised error. We will need to update these to make sure the plotting worked.

[dschwoerer]

I think there are several options:

1. remove the offending tests (simple)
2. reintroduce _ensure_plottable but only check for multiindex. This has still the issue that it might at some point be possible, and the tests need to be changed and _ensure_plottable needs to be changed
3. Change the test to check for any error. This basically just means we are checking it isn't working, which I see only limited value in.

Which of these options would you prefer?

[dcherian]

The error for multiindex plotting is not very informative "setting array element with sequence" so I would still keep _ensure_plottable to raise a nice error for that.

for the tests that now work, we should check that the output is reasonable, so I wouldn't delete them

[Illviljan]

What would you expect the plot to look like if you input a complex number or a multiindex along the x-axis?
I think I would expect it to be shown like a flattened array. And you can kind of get away with that if you cast the arrays to string before hand.

x = 1j * np.arange(0, 5)
y = np.arange(5, 10)
plt.plot(x, y)
ComplexWarning: Casting complex values to real discards the imaginary part
  return np.asarray(x, float)

x = 1j * np.arange(0, 5)
y = np.arange(5, 10)
plt.plot(np.vectorize(str)(x), y)

x = 1j + np.arange(0, 5)
y = np.arange(5, 10)
plt.plot(np.vectorize(str)(x), y)

import pandas as pd
arrays = [[1, 1, 2, 2], ['red', 'blue', 'red', 'blue']]
a = pd.MultiIndex.from_arrays(arrays, names=('number', 'color')).to_numpy()
b = np.arange(4, 8)
plt.plot(np.vectorize(str)(a), b)