Plotting grids across the subplots Python matplotlib

Question

I have tried the following:

d = [1,2,3,4,5,6,7,8,9]
f = [0,1,0,0,1,0,1,1,0]
fig = plt.figure()
fig.set_size_inches(30,10)
ax1 = fig.add_subplot(211)
line1 = ax1.plot(d,marker='.',color='b',label="1 row")
ax2 = fig.add_subplot(212)
line1 = ax2.plot(f,marker='.',color='b',label="1 row")
ax1.grid()
ax2.grid()
plt.show()

I got the following output :

But I was expecting the following output:

How I can get the grids across the two plots?

Answer 1

There is no built-in option to create inter-subplot grids. In this case I'd say an easy option is to create a third axes in the background with the same grid in x direction, such that the gridline can be seen in between the two subplots.

import matplotlib.pyplot as plt

d = [1,2,3,4,5,6,7,8,9]
f = [0,1,0,0,1,0,1,1,0]

fig, (ax1,ax2) = plt.subplots(nrows=2, sharex=True)
ax3 = fig.add_subplot(111, zorder=-1)
for _, spine in ax3.spines.items():
    spine.set_visible(False)
ax3.tick_params(labelleft=False, labelbottom=False, left=False, right=False )
ax3.get_shared_x_axes().join(ax3,ax1)
ax3.grid(axis="x")


line1 = ax1.plot(d, marker='.', color='b', label="1 row")
line1 = ax2.plot(f, marker='.', color='b', label="1 row")
ax1.grid()
ax2.grid()
plt.show()

Answer 2

Here is my solution:

import matplotlib.pyplot as plt

x1 = [1,2,3,4,5,6,7,8,9]
x2= [0,1,0,0,1,0,1,1,0]
x3= range(-10,0)
# frameon=False removes frames
# fig, (ax1,ax2, ax3) = plt.subplots(nrows=3, sharex=True, subplot_kw=dict(frameon=False))
fig, (ax1,ax2, ax3) = plt.subplots(nrows=3, sharex=True)
# remove vertical gap between subplots
plt.subplots_adjust(hspace=.0)

ax1.grid()
ax2.grid()
ax3.grid()

ax1.plot(x1)
ax2.plot(x2)
ax3.plot(x3)

Without frames subplot_kw=dict(frameon=False):

Answer 3

An option is to create a single plot then just offset the data. So one set plots above the other.

Answer 4

Solution

You can extend the grid lines by passing the appropriate coordinates through to Line2D with ax1.grid(ydata=[-space_between_axes, 1], clip_on=False).

import matplotlib.pyplot as plt

d = [1,2,3,4,5,6,7,8,9]
f = [0,1,0,0,1,0,1,1,0]
fig = plt.figure()
fig.set_size_inches(30,10)
ax1 = fig.add_subplot(211)
line1 = ax1.plot(d,marker='.',color='b',label="1 row")
ax2 = fig.add_subplot(212)
line1 = ax2.plot(f,marker='.',color='b',label="1 row")
plt.setp(ax1.get_xticklabels(), backgroundcolor="white")  # Eliminate strike-through.
space_between_plots = plt.rcParams["figure.subplot.hspace"]  # <==========
ax1.xaxis.grid(ydata=[-space_between_plots, 1], clip_on=False)  # <==========
ax1.yaxis.grid()
ax2.grid()
plt.show()

How does this work?

• Per the docs of the matplotlib.axis.Axis.grid == ax1.xaxis.grid method, keyword arguments can be passed on to Line2D.
• Line2D takes the x- and y-coordinates (xdata and ydata) as arguments.
• But what coordinate system is used? You can check this with ax1.get_xgridlines()[-1].get_xydata(). The output is array([[9., 0.], [9., 1.]]). Apparently the x-values are in Data coordinates and the y-values are in Axes coordinates.
• Next question: How far down do you need to extend? In this particular case you can just take the number -1 because the overlap with the same lines further down is invisible. The proper way however is to get the actual distance, in this case with space_between_plots = plt.rcParams["figure.subplot.hspace"]. That's 0.2 which means one fifth of the height of one plot.
• The top stays the same at 1.
• clip_on=True is needed because Matplotlib doesn't draw anything outside the plotting area by default.

Note

It looks like Matplotlib has a bug that makes this solution unworkable in some circumstances. E.g. if I leave out plt.setp(...), Maptlotlib complains that it receives the ydata argument multiple times and raises a TypeError.