Shared drawing area problem

One . Preface

Sometimes we will find that if the range of the axis is not appropriate, it is difficult to extract the corresponding information from the picture , For example, the figure on the left and the figure below, it is obvious that we can't draw the difference between the density of the two . In addition, many times we want to show two different drawings under a coordinate axis system , But the axis fields of the two are different , What to do at this time ？

---

Let's take another example , Let's say we want to draw a on a coordinate axis sin Function and a cos function , Obviously, the following program can meet our needs .

fig, ax = plt.subplots(figsize=(10, 8))
x = np.linspace(0, 2*np.pi, 100)
ax.plot(np.sin(x), lw=2, color='c', label=r'$\sin({x})$')
ax.plot(np.cos(x), lw=2, color='k', label=r'$\cos({x})$')
ax.axhline(0, lw=4, color='r')
plt.grid(color='gray', ls=":")
plt.legend(loc='upper right', title=r'$\sin({x})$ and $\cos({x})$ Example ', title_fontsize=12, shadow=True, fancybox=True, framealpha=0.8)
plt.show()

The drawing results are as follows ：

Next , Suppose we use the same method to draw 20*sin(x) and cos(x) So what happens ？ Please look at the chart below.

We can see that in order to display the complete sinusoidal function normally , The entire cosine function has been compressed , Obviously, this will not achieve the effect we want . But we can see from observation , The definition domain of the two is the same , Only the value range is different , In other words, the two can share one X Axis , But you can't use the same Y Axis , This is the problem we need to solve , It is also the practical significance of sharing the coordinate axis of a single region . This paper mainly discusses the solutions to these two kinds of problems , namely Share the coordinate axis of a single area and Share the axes of the drawing areas of different subareas . The code in this article only provides the key part of the code , The import of necessary libraries and the initialization settings of libraries have been omitted .

---

Two . Solution

2.1 Axis solutions that share a single region

Through the above example, we already know that the core of solving this problem is how to share X Axis without sharing Y Axis . We can use twinx() Method to create and return a shared X The second axis system of the axis . Here's the definition of the method ：

twinx(ax=None)

Parameters 1：ax： Specify which axis system to share
It should be noted here that the new axis system will cover the old axis system , Its Y Generate ridge labels at the scale axis and scale line . For the question just now, the code using this method is as follows ：

fig, ax = plt.subplots(figsize=(10, 8))
x = np.linspace(0, 2*np.pi, 100)

line1 = ax.plot(20 * np.sin(x), lw=2, color='c', label=r'$\sin({x})$')
ax.axhline(0, lw=4, color='r')

ax2 = ax.twinx()
line2 = ax2.plot(np.cos(x), lw=2, color='k', label=r'$\cos({x})$')

plt.grid(color='gray', ls=":")
fig.suptitle(r'$\sin({x})$ and $\cos({x})$ Example ', fontsize=20)
plt.show()

The drawing results are as follows ：

You can see that the two share X But each has its own Y Axis , The larger axis will shrink to meet the smaller axis . In this way, we can also draw in a single area in the face of two maps with large data gap .

---

2.1 Axis solutions that share drawing areas of different subareas

There are two solutions for different drawing areas sharing different subareas , The first is to use set_xlim() and set_ylim() Method to manually modify the coordinate axis range , Of course, modifying shared areas is just the tip of the iceberg of these two methods , It doesn't prevent us from achieving the effect we want . The sample code is as follows ：

x1 = np.linspace(-2*np.pi, 2*np.pi, 500)
y1 = np.cos(x1**2)

x2 = np.linspace(0.01, 10, 100)
y2 = np.sin(x2)

x3 = np.random.randint(0, 10, 50)
y3 = np.linspace(0, 3, 50)

x4 = np.random.randint(0, 25, 50)
y4 = np.linspace(0, 20, 50)

fig, ax = plt.subplots(1, 2)
fig.suptitle(' Use set_xlim and set_ylim Method ', fontsize=25)
ax[0].scatter(x3, y3)
ax[0].set_title(" Modify the scope according to the figure on the right ", fontsize=15)
ax[1].scatter(x4, y4)
ax[1].set_title('X Axis value :[-2, 25],Y Axis value :[-2, 20]', fontsize=15)
plt.subplots_adjust(wspace=0.6, top=0.8)

ax[0].set_xlim((-2, 25))
ax[0].set_ylim((-2, 20))

ax[1].set_xlim((-2, 25))
ax[1].set_ylim((-2, 20))
plt.show()

The drawing results are as follows ：

From the above figure, we can easily see which of the two is more dense , Compared with the previous error examples, the visual effect of this time has been greatly improved .

---

The second solution is to specify the sharing mechanism at the beginning , It's about delivering sharex and sharey Parameters to subplots Method . The sample code is as follows ：

x1 = np.linspace(-2*np.pi, 2*np.pi, 500)
y1 = np.cos(x1**2)

x2 = np.linspace(0.01, 10, 100)
y2 = np.sin(x2)

x3 = np.random.randint(0, 10, 50)
y3 = np.linspace(0, 3, 50)

x4 = np.random.randint(0, 25, 50)
y4 = np.linspace(0, 20, 50)

fig, ax = plt.subplots(1, 2, sharey=True, sharex=True)
fig.suptitle(' Use set_xlim and set_ylim Method ', fontsize=25)
ax[0].scatter(x3, y3)
ax[0].set_title(" Modify the scope according to the figure on the right ", fontsize=15)
ax[1].scatter(x4, y4)
ax[1].set_title('X Axis value :[-2, 25],Y Axis value :[-2, 20]', fontsize=15)
plt.subplots_adjust(wspace=0.6, top=0.8)

plt.show()

The drawing results are as follows ：

It can be seen that the effects of these two methods are consistent . These two parameters have a total of five valid values , Namely ：

1. True/‘all’： All subareas share the same coordinate system
2. ‘col’： Each column in the specified sub area wants to have the same coordinate system
3. ‘row’： Each row in the specified sub area wants to have the same coordinate system
4. False/‘none’： Do not share
5. ax： Specify a sub area , In this way, any two sub areas can be specified separately to share the coordinate axis system

Be sure to remember that the shared coordinate system is the maximum of all subareas . That is, there may be X Axis and Y When the axis does not come from a sub area .

---

3、 ... and . Reference resources

1. Visualization solutions
2. The road to Visualization