How can I rotate a matplotlib plot through 90 degr

问题:

I have created a figure in matplotlib which contains three subplots, one in the top left quadrant, one in the top right quadrant, and one in the bottom right quadrant. The top right figure contains a two-d image, and the other two plots are the projection onto the Y and X axis respectively. I'd like to rotate the top left quadrant subplot through 90deg counterclockwise, so that the x-axis of that plot lies along the y-axis of the 2-d plot.

For the subplot, I realize I could flip the x and y data, rotate the axis labels, create a plot title on the left hand side, etc. But I was hoping to find a single call which would just rotate the whole, finished plot through 90deg. But I can't find one.

Is there a simple way to do this?

回答1:

Many of the pyplot 1D plots seem to have "orientation" or "pivot" options within their own arguments. For example, from matplotlib.org example of histogram:

matplotlib.pyplot.hist(x, 
                       bins=10, 
                       range=None, 
                       normed=False, 
                       weights=None, 
                       cumulative=False, 
                       bottom=None, 
                       histtype=u'bar', 
                       align=u'mid', 
                       orientation=u'vertical', 
                       rwidth=None, 
                       log=False, 
                       color=None, 
                       label=None, 
                       stacked=False, 
                       hold=None, 
                       **kwargs)

Just change to horizontal (orientation=u'vertical')

回答2:

Another interesting parameter for a lot of functions is transform (unlike orientation or pivot this parameter can also be used in e.g. plot).

The transform parameter allows you to add a transformation, specified by a Transform object. For the sake of example, this is how you would rotate the plot of some random data:

import numpy
from matplotlib import pyplot, transforms

data = numpy.random.randn(100)

# first of all, the base transformation of the data points is needed
base = pyplot.gca().transData
rot = transforms.Affine2D().rotate_deg(90)

# define transformed line
line = pyplot.plot(data, 'r--', transform= rot + base)
# or alternatively, use:
# line.set_transform(rot + base)

pyplot.show()

For an example on how to rotate a patch, see this answer, which was also the source of inspiration for this answer.

---

update

I recently found out that the transform parameter does not work as expected when using pyplot.scatter (and other PathCollections). In this case, you might want to use the offset_transform. See this answer for more information on how to the offset_transform can be set.

回答3:

The easiest way I could imagine would be to use the scipy rotate method on the figure. However, this requires the Python Imaging Library, which is not available for Python 3.x.

import scipy
from scipy import ndimage
import matplotlib.pyplot as plt

Your_Plot = plt.plot(X,Y)
Rotated_Plot = ndimage.rotate(Your_Plot, 90)

plt.figure(figsize=(12.5, 2.5))

plt.subplot(2,2,1)
plt.imshow(Rotated_Plot, cmap=plt.cm.gray)
plt.axis('off')

plt.show()

That's not a very nice way to do it though, I think a wrapper function or just manually swapping X and Y data, rotating axis labels, etc. would be far easier.

回答4:

from mpl_toolkits.mplot3d import axes3d
import matplotlib.pyplot as plt

fig=plt.figure() 
ax=fig.add_subplot(111,projection='3d')

# for rotate the axes and update.
for angle in range(0,360): 
    ax.view_init(30,angle)

plt.show()