Set default color for only one level contour

[zxdawn]

Description

contour doesn't work when only one level is available in levels.

Steps to reproduce

import proplot as plot
import pandas as pd
import numpy as np
fig, axs = plot.subplots()
state = np.random.RandomState(51423)
data = state.rand(6, 6)
data = pd.DataFrame(data, index=pd.Index(['a', 'b', 'c', 'd', 'e', 'f']))

# Line contours with labels
axs.contour(
    data.cumsum(axis=1) - 2,
    # color='gray8',
    levels=[0]
)
axs.format(title='Line contour plot with labels')

Expected behavior: Just like the color is specified:

color='gray8'

Actual behavior:

Traceback (most recent call last):
  File "/home/xin/Desktop/test.py", line 13, in <module>
    levels=[0]
  File "/home/xin/Software/miniconda3/envs/s5p/lib/python3.7/site-packages/proplot/ui.py", line 682, in _iterator
    result.append(func(*args, **kwargs))
  File "/home/xin/Software/miniconda3/envs/s5p/lib/python3.7/site-packages/proplot/axes/plot.py", line 3867, in wrapper
    return driver(self, func, *args, **kwargs)
  File "/home/xin/Software/miniconda3/envs/s5p/lib/python3.7/site-packages/proplot/axes/plot.py", line 897, in standardize_2d
    return func(self, x, y, *Zs, **kwargs)
  File "/home/xin/Software/miniconda3/envs/s5p/lib/python3.7/site-packages/proplot/axes/plot.py", line 3867, in wrapper
    return driver(self, func, *args, **kwargs)
  File "/home/xin/Software/miniconda3/envs/s5p/lib/python3.7/site-packages/proplot/internals/warnings.py", line 97, in wrapper
    return func_orig(*args, **kwargs)
  File "/home/xin/Software/miniconda3/envs/s5p/lib/python3.7/site-packages/proplot/axes/plot.py", line 2867, in cmap_changer
    minlength=(1 if name in ('contour', 'tricontour') else 2),
  File "/home/xin/Software/miniconda3/envs/s5p/lib/python3.7/site-packages/proplot/axes/plot.py", line 2552, in _build_discrete_norm
    levels, descending = pcolors._check_levels(levels)
  File "/home/xin/Software/miniconda3/envs/s5p/lib/python3.7/site-packages/proplot/colors.py", line 1760, in _check_levels
    raise ValueError(f'Levels {levels} must be 1d vector with size >= 2.')
ValueError: Levels [0] must be 1d vector with size >= 2.

Proplot version

0.6.3

[lukelbd]

This was due to the fact that DiscreteNorm (used for all colormaps in proplot) requires at least 2 levels. I added an edge condition to fix it -- the result of your example is now a single contour whose color is the central color of the specified (or default) colormap, which matches the behavior in matplotlib.

Example:

import proplot as plot
import pandas as pd
import numpy as np
fig, axs = plot.subplots()
state = np.random.RandomState(51423)
data = state.rand(6, 6)
data = pd.DataFrame(data, index=pd.Index(['a', 'b', 'c', 'd', 'e', 'f']))

# Line contours with labels
axs.contour(
    data.cumsum(axis=1) - 2,
    cmap='grays',
    levels=[0],
)
axs.format(title='Line contour plot with labels')

[zxdawn]

Thanks Luke! Nice fix ;)