I'm not a great fan of matplotlib despite how central it's become in data science. This is because of its API: I find it difficult to learn which suggests to me that it's counterintuitive. As long as I'm working with it I can do a lot but the moment I take a break to work on something else it's as if it all evaporates.
For this reason I've put together a barebones outline that I can come to when speed is of the essence.
import numpy as np import matplotlib.pyplot as plt # aspect ratio figsize = plt.figaspect( 2. ) fig = plt.figure( figsize=figsize ) # figure annotations fig.set_title( "..." ) fig.suptitle( "..." ) ax.set_ylabel( "..." ) # subplot ax = fig.add_subplot( 2, 1, 1 ) fig, ax = plt.subplots() fig, (ax0, ax1 ) = plt.subplots( nrows=2, sharex=True ) # lineplot l = ax.plot( ... ) l.set_dashes( ) # surface plot ax = fig.add_subplot( 2, 1, 2, projection='3d' ) surf = ax.plot.surface( ... ) ax.set_zlim3d( -1, 1 ) # histogram/barplot n, bins, patches = plt.hist( ... ) rects = ax.bar( ... ) autolabel( rects ) # add a grid, modify axes lines and ticks ax.grid( True ) ax.set_axis_off() ax.spines['right'].set_visible( False ) ax.spines['right].set_position(( 'outward', 10 )) ax.yaxis.set_tick_position( 'left' ) ax.set_xlim(( 0, 2*np.pi )) ax.set_xticks([ 0, np.pi, 2*np.pi ]) ax.set_ticklabels([ '0', '$\pi$', '2$\pi$' ]) # display plt.subplots_adjust( hspace=0.3 ) plt.margins( 0.2 ) plt.show()