如何在Seaborn Jointplot中对单个标记进行着色？

Question

I believe seaborn wants RGBA color format so I gave it a list of RGBA tuples like in matplotlib but it didn't like it. I tried doing something like g = sns.jointplot(plotdataandstuff) but I couldn't figure out how to navigate the <seaborn.axisgrid.JointGrid at 0x135943d30> object

u = np.array([ 3.41959, 1.79315, 1.17229, 1.59909, 1.27337, 1.21917, 2.60591, 2.0571 , 1.83865, 1.94869])
v = np.array([238614, 166714, 152668, 131310, 131249, 130562, 118618, 117329, 101841, 87626])
c = np.array([(0.38998137426811552, 0.61096214975059782, 0.40096663322878834, 1.0), 
              (0.91819597903385741, 0.79247826098536023, 0.76754610185633376, 1.0), 
              (0.80693280982205806, 0.72756928157840872, 0.59917426429032361, 1.0), 
              (0.67210067882600855, 0.69086449405289441, 0.48640673089730696, 1.0), 
              (0.57590730770024134, 0.66895624366124595, 0.43845220910273563, 1.0), 
              (0.63252315055536124, 0.68193477278802606, 0.4639878928469835, 1.0), (
                  0.72522872317321141, 0.70350066480196072, 0.52329091024506846, 1.0), 
              (0.67210067882600855, 0.69086449405289441, 0.48640673089730696, 1.0), 
              (0.93684201381384091, 0.81480601869516889, 0.81083911715694279, 1.0), 
              (0.83654311363401856, 0.73919037695009648, 0.63434086709274395, 1.0)], dtype=object)
sns.jointplot(x=u, y=v, color=c)

Here's the error

TypeError                                 Traceback (most recent call last)
/Users/jespinoz/anaconda/lib/python3.5/site-packages/matplotlib/colors.py in to_rgba(c, alpha)
    140     try:
--> 141         rgba = _colors_full_map.cache[c, alpha]
    142     except (KeyError, TypeError):  # Not in cache, or unhashable.

TypeError: unhashable type: 'numpy.ndarray'

During handling of the above exception, another exception occurred:

TypeError                                 Traceback (most recent call last)
/Users/jespinoz/anaconda/lib/python3.5/site-packages/matplotlib/colors.py in _to_rgba_no_colorcycle(c, alpha)
    191     try:
--> 192         c = tuple(map(float, c))
    193     except TypeError:

TypeError: only length-1 arrays can be converted to Python scalars

During handling of the above exception, another exception occurred:

ValueError                                Traceback (most recent call last)
<ipython-input-69-48b550e1edb6> in <module>()
     13               (0.93684201381384091, 0.81480601869516889, 0.81083911715694279, 1.0),
     14               (0.83654311363401856, 0.73919037695009648, 0.63434086709274395, 1.0)], dtype=object)
---> 15 sns.jointplot(x=u, y=v, color=c)

/Users/jespinoz/anaconda/lib/python3.5/site-packages/seaborn/distributions.py in jointplot(x, y, data, kind, stat_func, color, size, ratio, space, dropna, xlim, ylim, joint_kws, marginal_kws, annot_kws, **kwargs)
    789     if color is None:
    790         color = color_palette()[0]
--> 791     color_rgb = mpl.colors.colorConverter.to_rgb(color)
    792     colors = [set_hls_values(color_rgb, l=l) for l in np.linspace(1, 0, 12)]
    793     cmap = blend_palette(colors, as_cmap=True)

/Users/jespinoz/anaconda/lib/python3.5/site-packages/matplotlib/colors.py in to_rgb(arg)
    307         if *arg* is *RGBA*, the *A* will simply be discarded.
    308         """
--> 309         return to_rgb(arg)
    310 
    311     @staticmethod

/Users/jespinoz/anaconda/lib/python3.5/site-packages/matplotlib/colors.py in to_rgb(c)
    242     """Convert `c` to an RGB color, silently dropping the alpha channel.
    243     """
--> 244     return to_rgba(c)[:3]
    245 
    246 

/Users/jespinoz/anaconda/lib/python3.5/site-packages/matplotlib/colors.py in to_rgba(c, alpha)
    141         rgba = _colors_full_map.cache[c, alpha]
    142     except (KeyError, TypeError):  # Not in cache, or unhashable.
--> 143         rgba = _to_rgba_no_colorcycle(c, alpha)
    144         try:
    145             _colors_full_map.cache[c, alpha] = rgba

/Users/jespinoz/anaconda/lib/python3.5/site-packages/matplotlib/colors.py in _to_rgba_no_colorcycle(c, alpha)
    192         c = tuple(map(float, c))
    193     except TypeError:
--> 194         raise ValueError("Invalid RGBA argument: {!r}".format(orig_c))
    195     if len(c) not in [3, 4]:
    196         raise ValueError("RGBA sequence should have length 3 or 4")

ValueError: Invalid RGBA argument: array([[0.3899813742681155, 0.6109621497505978, 0.40096663322878834, 1.0],
       [0.9181959790338574, 0.7924782609853602, 0.7675461018563338, 1.0],
       [0.8069328098220581, 0.7275692815784087, 0.5991742642903236, 1.0],
       [0.6721006788260085, 0.6908644940528944, 0.48640673089730696, 1.0],
       [0.5759073077002413, 0.668956243661246, 0.43845220910273563, 1.0],
       [0.6325231505553612, 0.6819347727880261, 0.4639878928469835, 1.0],
       [0.7252287231732114, 0.7035006648019607, 0.5232909102450685, 1.0],
       [0.6721006788260085, 0.6908644940528944, 0.48640673089730696, 1.0],
       [0.9368420138138409, 0.8148060186951689, 0.8108391171569428, 1.0],
       [0.8365431136340186, 0.7391903769500965, 0.634340867092744, 1.0]], dtype=object)

Answer 1

sns.jointplot(u, v, joint_kws={'color':None, 'c':c.tolist()})

A joinplot consist of the marginal plots on the sides (histograms by default) and the jointplot in the middle (a scatterplot by default), which you can read more about in the JointGrid documentation .

In this case, you want to pass arguments to the scatter plot in the middle, which you can do with the joint_kws parameter in sns.jointplot() . This parameter takes a dictionary of parameter names and their arguments, and passes it to the function used to create the middle plot, in this case plt.scatter() . The c parameter controls the colors, and accepts a list of color tuples, so you will need to convert the array via c.tolist() . One last caveat is that seaborn already passes a color to the color parameter of plt.scatter() , so we need to set this to None to avoid raising an error from trying to set the marker colors twice.