python - matplotlib savefig image size with bbox_inches='tight'

Question

I have to make a vector plot and I want to just see the vectors without the axes, titles etc so here is how I try to do it:

pyplot.figure(None, figsize=(10, 16), dpi=100)
pyplot.quiver(data['x'], data['y'], data['u'], data['v'], 
              pivot='tail', 
              units='dots', 
              scale=0.2,
              color='black')

pyplot.autoscale(tight=True)
pyplot.axis('off')
ax = pyplot.gca()
ax.xaxis.set_major_locator(pylab.NullLocator())
ax.yaxis.set_major_locator(pylab.NullLocator())
pyplot.savefig("test.png", 
               bbox_inches='tight', 
               transparent=True,
               pad_inches=0)

and despite my efforts to have an image 1000 by 1600 I get one 775 by 1280. How do I make it the desired size? Thank you.

UPDATE The presented solution works, except in my case I also had to manually set the axes limits. Otherwise, matplotlib could not figure out the "tight" bounding box.

With MPL, there are two DPI values you have to keep straight. The one that you specified when you created the figure object is for the interactive display of the figure on your screen. The other DPI value taht applies to the saved file (in whatever format) is specified when you call savefig. That is why the solution @unutbu posted works. — Paul H

unutbu unutbu · Accepted Answer · 2012-10-22T19:54:38

import matplotlib.pyplot as plt
import numpy as np
sin, cos = np.sin, np.cos

fig = plt.figure(frameon = False)
fig.set_size_inches(5, 8)
ax = plt.Axes(fig, [0., 0., 1., 1.], )
ax.set_axis_off()
fig.add_axes(ax)

x = np.linspace(-4, 4, 20)
y = np.linspace(-4, 4, 20)
X, Y = np.meshgrid(x, y)
deg = np.arctan(Y**3-3*Y-X)
plt.quiver(X, Y, cos(deg), sin(deg), pivot = 'tail', units = 'dots', color = 'red', )
plt.savefig('/tmp/test.png', dpi = 200)

yields

enter image description here

You can make the resultant image 1000x1600 pixels by setting the figure to be 5x8 inches

fig.set_size_inches(5, 8)

and saving with DPI = 200:

plt.savefig('/tmp/test.png', dpi = 200)

The code to remove the border was taken from here.

(The image posted above is not to scale since 1000x1600 is rather large).

python - matplotlib savefig image size with bbox_inches='tight'

1 Answers