Bug summary

When creating a LogNorm imshow plot with more than 9 orders of magnitude of dynamic range the automatically generated colorbar from plt.colorbar() has no automatically generated tick marks or labels.

Code for reproduction

This generates a plot with colorbar tick labels

importnumpyasnpfrommatplotlibimportpyplotaspltfrommatplotlib.colorsimportLogNormdata=np.zeros((800,800))+1data[50,51]=1e9print(np.log10(data.max()/data.min()))image=plt.imshow(data,origin='lower',interpolation='nearest',norm=LogNorm())plt.colorbar(image)plt.savefig('labels.png')

While this script generates a plot with no colorbar tick labels:

importnumpyasnpfrommatplotlibimportpyplotaspltfrommatplotlib.colorsimportLogNormdata=np.zeros((800,800))+1data[50,51]=2e9print(np.log10(data.max()/data.min()))image=plt.imshow(data,origin='lower',interpolation='nearest',norm=LogNorm())plt.colorbar(image)plt.savefig('no-labels.png')

Expected outcome

I would expect the second plot to also have colorbar tick labels.

Matplotlib version

• Operating system: Mac OS High Sierra
• Matplotlib version: 2.1.1 (installed via pip)
• Matplotlib backend (print(matplotlib.get_backend())): MacOSX
• Python version: 3.6