[FIXED] Matplotlib plot shows 2 labels on the y-axis

Issue

I am trying to generate a continously generated plot in matplotlib. The problem that I am facing is related to the labelling on the right y-axis. The range that shows is my desired, however there is also an additional set off labels (0, 0.2, ... 1,0).

def doAnimation():
    fig, ax = plt.subplots()

    def animate(i):
        data=prices(a,b,c)  #this gives a DataFrame with 2 columns (value 1 and 2)

        plt.cla()
        ax.plot(data.index, data.value1)
        ax2 = ax.twinx()
        ax2.plot(data.index, data.value2)
        plt.gcf().autofmt_xdate()     
        plt.tight_layout()  
        return ax, ax2

    call = FuncAnimation(plt.gcf(), animate, 1000)  
    return call

callSave = doAnimation()
plt.show()

Any ideas how can I get rid of the set: 0.0, 0.2, 0.4, 0.6, 0.8, 1.0?

This is how the graph looks:

Solution

plt.cla clears the current axes. The first time you call plt.cla, the current axes are ax (ax2 doesn't exist yet). Clearing these axes resets both the x and y range of ax to (0,1). However, on line 8, you plot to ax, and both ranges are appropriately adjusted.

On line 9, you create a new set of axes and call them ax2. When you leave the animate function, the name ax2 will go out of scope, but the axes object to which it refers will persist. These axes are now the current axes.

The second time you call animate, plt.cla clears those axes, resetting the x and y range to (0,1). Then, on line 9, you create a new set of axes and call them ax2. These axes are not the same axes as before! ax2 in fact refers to a third set of axes, which will be cleared the next time you call plt.cla. Each new call to animate makes a new set of axes. The offending axes labels appear to be bolded; in fact, they have been drawn a thousand times.

The simplest (fewest changes) fix would be to move ax2 = ax.twinx() outside of animate, and replace plt.cla with separate calls to ax.cla and ax2.cla.

I think a better approach would be to create the lines outside of animate, and modify their data within animate. While we're at it, let's replace those references to plt.gcf() with references to fig, and set tight_layout via an argument to plt.subplots.

Putting said changes together, we get,

import matplotlib.pyplot as plt
from matplotlib.animation import FuncAnimation
import pandas as pd
import numpy as np


def dummy_prices():
    samples = 101
    xs = np.linspace(0, 10, samples)
    ys = np.random.randn(samples)
    zs = np.random.randn(samples) * 10 + 50
    return pd.DataFrame.from_records({'value1': ys, 'value2': zs}, index=xs)


def doAnimation():
    fig, ax = plt.subplots(1, 1, tight_layout=True)
    fig.autofmt_xdate()
    ax2 = ax.twinx()

    data = dummy_prices()
    line = ax.plot(data.index, data.value1)[0]
    line2 = ax2.plot(data.index, data.value2, 'r')[0]

    def animate(i):
        data = dummy_prices()
        line.set_data(data.index, data.value1)
        line2.set_data(data.index, data.value2)
        return line, line2

    animator = FuncAnimation(fig, animate, frames=10)
    return animator


def main():
    animator = doAnimation()
    animator.save('animation.gif')


if __name__ == '__main__':
    main()

where animation.gif should look something like,

Answered By - fanjie