使用LogNorm和ImageGrid时的Matplotlib颜色条

Question

I have been trying to plot some panels using ImageGrid . When I use grid.cbar_axes[0].colorbar(im) to set the colorbar the colors look fine, but the scale on the colorbar reads like it's going from 2x10^0 to None .

I have tried dozens of workarounds but nothing worked. Here's the figure I'm trying to make (my wrong version):

Unfortunately I couldn't make a MWE that perfectly reproduces the problem. I did produce a MWE that partially reproduces it. If I use this code:

import matplotlib.pyplot as plt
from mpl_toolkits.axes_grid1 import ImageGrid
from matplotlib.colors import LogNorm

def get_demo_image():
    import numpy as np
    from matplotlib.cbook import get_sample_data
    f = get_sample_data("axes_grid/bivariate_normal.npy", asfileobj=False)
    z = np.load(f)
    return abs(z), (-3, 4, -4, 3)

F = plt.figure(figsize=(5.5, 3.5))
grid = ImageGrid(F, 111,  # similar to subplot(111)
                 nrows_ncols=(1, 3),
                 axes_pad=0.1,
                 add_all=True,
                 label_mode="L",
                 cbar_mode='single'
                 )

Z, extent = get_demo_image()  # demo image

im1 = Z
im2 = Z
im3 = Z
vmin, vmax = 1e-3, 1e10
for i, im in enumerate([im1, im2, im3]):
    ax = grid[i]
    imc = ax.pcolormesh(range(15), range(15), im, norm=LogNorm(vmin=vmin, vmax=vmax), linewidth=0, rasterized=True)

cb=grid.cbar_axes[0].colorbar(imc)

I get almost the same behavior, except that the upper limit appears to be fine. The lower limit still presents the same weird behavior, no matter what values I use for vmin and vmax .

Any idea of what might be going on?

Answer 1

Although the official example uses something similar to cb=grid.cbar_axes[0].colorbar(imc) , which would translate here into grid[2].cax.colorbar(im) , I'm lost on why that would even make sense.

Instead, the usual way to produce a colorbar would also work here, using the colorbar method of Figure with the ScalarMappable (here imc ) as argument and the axes to produce the colorbar as keywordargument to cax (here grid[2].cax ):

cb=F.colorbar(imc, cax=grid[2].cax)