How to sharex when using subplot2grid
Question:
I’m a Matlab user recently converted to Python. Most of the Python skills I manage on my own, but with plotting I have hit the wall and need some help.
This is what I’m trying to do…
I need to make a figure that consists of 3 subplots with following properties:
- subplot layout is 311, 312, 313
- the height of 312 and 313 is approximately half of the 311
- all subplots share common X axis
- the space between the subplots is 0 (they touch each other at X axis)
By the way I know how to make all this, only not in a single figure. That is the problem I’m facing now.
For example, this is my ideal subplot layout:
import numpy as np
import matplotlib.pyplot as plt
t = np.arange(0.0, 2.0, 0.01)
s1 = np.sin(2*np.pi*t)
s2 = np.exp(-t)
s3 = s1*s2
fig = plt.figure()
ax1 = plt.subplot2grid((4,3), (0,0), colspan=3, rowspan=2)
ax2 = plt.subplot2grid((4,3), (2,0), colspan=3)
ax3 = plt.subplot2grid((4,3), (3,0), colspan=3)
ax1.plot(t,s1)
ax2.plot(t[:150],s2[:150])
ax3.plot(t[30:],s3[30:])
plt.tight_layout()
plt.show()
Notice how the x axis of different subplots is misaligned. I do not know how to align the x axis in this figure, but if I do something like this:
import numpy as np
import matplotlib.pyplot as plt
t = np.arange(0.0, 2.0, 0.01)
s1 = np.sin(2*np.pi*t)
s2 = np.exp(-t)
s3 = s1*s2
fig2, (ax1, ax2, ax3) = plt.subplots(nrows=3, ncols=1, sharex=True)
ax1.plot(t,s1)
ax2.plot(t[:150],s2[:150])
ax3.plot(t[30:],s3[30:])
plt.tight_layout()
plt.show()
Now the x axis is aligned between the subplots, but all subplots are the same size (which is not what I want)
Furthermore, I would like that the subplots are touching at x axis like this:
import numpy as np
import matplotlib.pyplot as plt
t = np.arange(0.0, 2.0, 0.01)
s1 = np.sin(2*np.pi*t)
s2 = np.exp(-t)
s3 = s1*s2
fig1 = plt.figure()
plt.subplots_adjust(hspace=0)
ax1 = plt.subplot(311)
ax2 = plt.subplot(312, sharex=ax1)
ax3 = plt.subplot(313, sharex=ax1)
ax1.plot(t,s1)
ax2.plot(t[:150],s2[:150])
ax3.plot(t[30:],s3[30:])
xticklabels = ax1.get_xticklabels()+ax2.get_xticklabels()
plt.setp(xticklabels, visible=False)
plt.show()
So to rephrase my question:
I would like to use
plt.subplot2grid(..., colspan=3, rowspan=2)
plt.subplots(..., sharex=True)
plt.subplots_adjust(hspace=0)
and
plt.tight_layout()
together in the same figure. How to do that?
Answers:
Just specify sharex=ax1 when creating your second and third subplots.
import numpy as np
import matplotlib.pyplot as plt
t = np.arange(0.0, 2.0, 0.01)
s1 = np.sin(2*np.pi*t)
s2 = np.exp(-t)
s3 = s1*s2
fig = plt.figure()
ax1 = plt.subplot2grid((4,3), (0,0), colspan=3, rowspan=2)
ax2 = plt.subplot2grid((4,3), (2,0), colspan=3, sharex=ax1)
ax3 = plt.subplot2grid((4,3), (3,0), colspan=3, sharex=ax1)
ax1.plot(t,s1)
ax2.plot(t[:150],s2[:150])
ax3.plot(t[30:],s3[30:])
fig.subplots_adjust(hspace=0)
for ax in [ax1, ax2]:
plt.setp(ax.get_xticklabels(), visible=False)
# The y-ticks will overlap with "hspace=0", so we'll hide the bottom tick
ax.set_yticks(ax.get_yticks()[1:])
plt.show()
If you still what to use fig.tight_layout(), you’ll need to call it before fig.subplots_adjust(hspace=0). The reason for this is that tight_layout works by automatically calculating parameters for subplots_adjust and then calling it, so if subplots_adjust is manually called first, anything in the first call to it will be overridden by tight_layout.
E.g.
fig.tight_layout()
fig.subplots_adjust(hspace=0)
A possible solution is to manually create the axis using the add_axis method like shown here:
import numpy as np
import matplotlib.pyplot as plt
t = np.arange(0.0, 2.0, 0.01)
s1 = np.sin(2*np.pi*t)
s2 = np.exp(-t)
s3 = s1*s2
left, width = 0.1, 0.8
rect1 = [left, 0.5, width, 0.4]
rect2 = [left, 0.3, width, 0.15]
rect3 = [left, 0.1, width, 0.15]
fig = plt.figure()
ax1 = fig.add_axes(rect1) #left, bottom, width, height
ax2 = fig.add_axes(rect2, sharex=ax1)
ax3 = fig.add_axes(rect3, sharex=ax1)
ax1.plot(t,s1)
ax2.plot(t[:150],s2[:150])
ax3.plot(t[30:],s3[30:])
# hide labels
for label1,label2 in zip(ax1.get_xticklabels(),ax2.get_xticklabels()):
label1.set_visible(False)
label2.set_visible(False)
plt.show()
But this way you cannot use tight_layout as you explicitly define the size of each axis.
I’m a Matlab user recently converted to Python. Most of the Python skills I manage on my own, but with plotting I have hit the wall and need some help.
This is what I’m trying to do…
I need to make a figure that consists of 3 subplots with following properties:
- subplot layout is 311, 312, 313
- the height of 312 and 313 is approximately half of the 311
- all subplots share common X axis
- the space between the subplots is 0 (they touch each other at X axis)
By the way I know how to make all this, only not in a single figure. That is the problem I’m facing now.
For example, this is my ideal subplot layout:
import numpy as np
import matplotlib.pyplot as plt
t = np.arange(0.0, 2.0, 0.01)
s1 = np.sin(2*np.pi*t)
s2 = np.exp(-t)
s3 = s1*s2
fig = plt.figure()
ax1 = plt.subplot2grid((4,3), (0,0), colspan=3, rowspan=2)
ax2 = plt.subplot2grid((4,3), (2,0), colspan=3)
ax3 = plt.subplot2grid((4,3), (3,0), colspan=3)
ax1.plot(t,s1)
ax2.plot(t[:150],s2[:150])
ax3.plot(t[30:],s3[30:])
plt.tight_layout()
plt.show()
Notice how the x axis of different subplots is misaligned. I do not know how to align the x axis in this figure, but if I do something like this:
import numpy as np
import matplotlib.pyplot as plt
t = np.arange(0.0, 2.0, 0.01)
s1 = np.sin(2*np.pi*t)
s2 = np.exp(-t)
s3 = s1*s2
fig2, (ax1, ax2, ax3) = plt.subplots(nrows=3, ncols=1, sharex=True)
ax1.plot(t,s1)
ax2.plot(t[:150],s2[:150])
ax3.plot(t[30:],s3[30:])
plt.tight_layout()
plt.show()
Now the x axis is aligned between the subplots, but all subplots are the same size (which is not what I want)
Furthermore, I would like that the subplots are touching at x axis like this:
import numpy as np
import matplotlib.pyplot as plt
t = np.arange(0.0, 2.0, 0.01)
s1 = np.sin(2*np.pi*t)
s2 = np.exp(-t)
s3 = s1*s2
fig1 = plt.figure()
plt.subplots_adjust(hspace=0)
ax1 = plt.subplot(311)
ax2 = plt.subplot(312, sharex=ax1)
ax3 = plt.subplot(313, sharex=ax1)
ax1.plot(t,s1)
ax2.plot(t[:150],s2[:150])
ax3.plot(t[30:],s3[30:])
xticklabels = ax1.get_xticklabels()+ax2.get_xticklabels()
plt.setp(xticklabels, visible=False)
plt.show()
So to rephrase my question:
I would like to use
plt.subplot2grid(..., colspan=3, rowspan=2)
plt.subplots(..., sharex=True)
plt.subplots_adjust(hspace=0)
and
plt.tight_layout()
together in the same figure. How to do that?
Just specify sharex=ax1 when creating your second and third subplots.
import numpy as np
import matplotlib.pyplot as plt
t = np.arange(0.0, 2.0, 0.01)
s1 = np.sin(2*np.pi*t)
s2 = np.exp(-t)
s3 = s1*s2
fig = plt.figure()
ax1 = plt.subplot2grid((4,3), (0,0), colspan=3, rowspan=2)
ax2 = plt.subplot2grid((4,3), (2,0), colspan=3, sharex=ax1)
ax3 = plt.subplot2grid((4,3), (3,0), colspan=3, sharex=ax1)
ax1.plot(t,s1)
ax2.plot(t[:150],s2[:150])
ax3.plot(t[30:],s3[30:])
fig.subplots_adjust(hspace=0)
for ax in [ax1, ax2]:
plt.setp(ax.get_xticklabels(), visible=False)
# The y-ticks will overlap with "hspace=0", so we'll hide the bottom tick
ax.set_yticks(ax.get_yticks()[1:])
plt.show()
If you still what to use fig.tight_layout(), you’ll need to call it before fig.subplots_adjust(hspace=0). The reason for this is that tight_layout works by automatically calculating parameters for subplots_adjust and then calling it, so if subplots_adjust is manually called first, anything in the first call to it will be overridden by tight_layout.
E.g.
fig.tight_layout()
fig.subplots_adjust(hspace=0)
A possible solution is to manually create the axis using the add_axis method like shown here:
import numpy as np
import matplotlib.pyplot as plt
t = np.arange(0.0, 2.0, 0.01)
s1 = np.sin(2*np.pi*t)
s2 = np.exp(-t)
s3 = s1*s2
left, width = 0.1, 0.8
rect1 = [left, 0.5, width, 0.4]
rect2 = [left, 0.3, width, 0.15]
rect3 = [left, 0.1, width, 0.15]
fig = plt.figure()
ax1 = fig.add_axes(rect1) #left, bottom, width, height
ax2 = fig.add_axes(rect2, sharex=ax1)
ax3 = fig.add_axes(rect3, sharex=ax1)
ax1.plot(t,s1)
ax2.plot(t[:150],s2[:150])
ax3.plot(t[30:],s3[30:])
# hide labels
for label1,label2 in zip(ax1.get_xticklabels(),ax2.get_xticklabels()):
label1.set_visible(False)
label2.set_visible(False)
plt.show()
But this way you cannot use tight_layout as you explicitly define the size of each axis.