[FIXED] How to appropriately set the limit of the Axes in matplotlib

Issue

I have a dataframe

import pandas as pd
import matplotlib.pyplot as plt
import numpy as np

data = {
    'id': [1, 2, 3, 4, 5,6,7,8,9,10],
    'LeftError': [0.1, 0.2, 0.15, 0.3, 0.25,-0.1, -0.2, -0.15, -0.3, -0.25],
    'RightError': [0.2, 0.3, 0.25, 0.4, 0.35,-0.2, -0.3, -0.25, -0.4, -0.35],
    'NCL': [1, 2, 1, 3, 2,-1, -2, -1, -3, -2],
    'NCR': [2, 3, 2, 4, 3,-2, -3, -2, -4, -3],
}

df = pd.DataFrame(data)

I want to plot it

    fig, axes = plt.subplots(1, 2, figsize=(12, 5))

    # Plot for Left side
    axes[0].scatter(df['NCL'], df['LeftError'], color='blue')

    axes[0].set_xlabel('NCL')
    axes[0].set_ylabel('Left Error')

    # Plot for Right side
    axes[1].scatter(df['NCR'], df['RightError'], color='green')

    axes[1].set_xlabel('NCR')
    axes[1].set_ylabel('Right Error')

    plt.show()

However when I do this I get

You can see that there is a good thing and a bad thing here. The good thing is that the Y axis goes a bit more and less than the range of values (it does not cross the points) . The bad thing is that they show different values. So to correct the bad thing I do

axes[0].set_ylim((min(df['LeftError'].min(), df['RightError'].min())), (max(df['LeftError'].max(), df['RightError'].max())))
axes[1].set_ylim((min(df['LeftError'].min(), df['RightError'].min())), (max(df['LeftError'].max(), df['RightError'].max())))

and now I get

So the bad thing was corrected, but the good thing became bad. See that the first and last green dot got crossed by the axis

How can I make the axis have some pad as originally but keeping the same values for both?

Solution

Instead of using the data, you can use the already-padded xlims and ylims of the plots. You just need to take the minimum and maximum of all the axes' limits.

import numpy as np
import matplotlib.pyplot as plt

plt.close("all")

x1 = np.linspace(-4, 4, 10)
x2 = np.linspace(-3, 3, 10)
y1 = 2*x1
y2 = 3*x2

fig, axes = plt.subplots(1, 2)
axes[0].scatter(x1, y1)
axes[1].scatter(x2, y2)

xlim = (min([ax.get_xlim()[0] for ax in axes]),
        max([ax.get_xlim()[1] for ax in axes]))
ylim = (min([ax.get_ylim()[0] for ax in axes]),
        max([ax.get_ylim()[1] for ax in axes]))

plt.setp(axes, xlim=xlim, ylim=ylim)

Answered By - jared