Can AxesGrid be used to plot two imshows (overlay) with two separate colorbars?

Question

I am using AxesGrid in matplotlib to create a plot that overlays two imshow plots, with a separate colourbar side by side for each image colourmap. While I can see in this question/answer that using pyplot.colorbar() automatically plots the second colour bar next to the first, this doesn't seem to work with AxesGrid.

    import matplotlib.pyplot as plt
    from mpl_toolkits.axes_grid1 import AxesGrid

    fig = plt.figure(1, facecolor='white',figsize=(10,7.5))

    grid = AxesGrid(fig, 111, 
                    nrows_ncols=(1, 1),
                    axes_pad=(0.45, 0.15),
                    label_mode="1",
                    share_all=True,
                    cbar_location="right",
                    cbar_mode="each",
                    cbar_size="7%",
                    cbar_pad="2%",
                    )

    im = grid[0].imshow(my_image, my_cmap)
    cbar = grid.cbar_axes[0].colorbar(im)

    im2 = grid[0].imshow(my_image_overlay, my_cmap2, alpha=0.5)
    cbar2 = grid.cbar_axes[0].colorbar(im2)

    plt.show()

However, this just shows the second colourbar. (Presumably overlaying the first one). I tried overriding the padding in cbar2 with:

    cbar2 = grid.cbar_axes[0].colorbar(im2, pad=0.5)

but this results in an error with "got an unexpected keyword argument 'pad'"

Is there a way to offset the second colourbar?

Answer 1

I think you may need to make two axes for the colorbars and use them:

from mpl_toolkits.axes_grid1 import make_axes_locatable
# create an axes on the right side of ax. The width of cax will be 5%
# of ax and the padding between cax and ax will be fixed at 0.05 inch.
divider = make_axes_locatable(ax)
cax = divider.append_axes("right", size="5%", pad=0.05)

'cause you're currently using the same allocated space for both.