Project Bivariate and Multivariate Colormapping, which resolves github.com//issues/14168 is divided into a number of milestones, of which this is one (highlighted in bold).

• Classes for bivariate and multivariate colormaps
• Colorizer and ColorizingArtist to replace ScalarMappable and handle the norm→color pipeline
• A MultiNorm class to handle the normalization of data for bivariate and multivariate colormaps
• Ensure the Colorizer and ColorizingArtist work with MultiNorm, BivarColormap and MultivarColormap
• Expose the new functionality via the top-level plotting functions: imshow(), pcolor() and pcolormesh()
• Create equivalents to fig.colorbar() for BivarColormap and MultivarColormap to work with ColorizingArtist
• Select bivariate and multivariate colormaps to include in matplotlib
• Examples demonstrating the new functionality

The following prototype uses a hypothetical function fig.colorbar_2D(ColorizingArtist) to display a BivarColormap:

x = np.linspace(-1.5, 0.5, 200)
y = np.linspace(-1, 1, 200)
xx, yy = np.meshgrid(x, y)
c = xx+1j*yy
z = c
for i in range(7):
    z = z**2+c

fig, ax = plt.subplots(1, 1, figsize=(10, 6))

cm = ax.pcolormesh(xx, yy, (z.imag, z.real), cmap='BiCone', vmin=-1, vmax=1)
ax.set_xlabel('Re{$c$}')
ax.set_ylabel('Im{$c$}')
ax.set_title('Mandelbrot $z_{7}$\n $z_i = z_{i-1}^2+c$')

cmap = fig.colorbar_2D(cm)
cmap.set_ylabel('Im{$z$}')
cmap.set_xlabel('Re{$z$}')

fig.show()

While the name colorbar_2D is provisional, this functionality will be required as part of Bivariate and Multivariate Colormapping.

Main questions include:

• Function name
• Should the function return an a new subclass of Axes? (subclass colorbar?) What members do the new class need?
• Placement

Similar questions arise for the MultivarColormap class, i.e. in making the following plot:

No work is expected on this issue until the milestones above it in Bivariate and Multivariate Colormapping have been reached.