Live Plotting

Jupyter Notebook is grate for exploration, but real systems run continuously. Let's learn how to plot live with matplotlib.

Make sure that you have Miniconda and VSCode installed, and follow the code below.

VSCode

import numpy as np
import matplotlib.pyplot as plt
import time

plt.ion() # interactive mode on

t_data = []
y_data = []

start = time.time()

while True:

    t = time.time() - start
    y = np.sin(2 * np.pi * 1 * t) + 0.8*np.random.random()

    t_data.append(t)
    y_data.append(y)

    # Filter noise
    #-----------------------------
    window = 5
    if(len(y_data) > window):
        y_smooth = np.convolve(y_data, np.ones(window)/window, mode='same')
    else:
        y_smooth = y_data
    #-----------------------------

    plt.clf() # clear the plot for the next frame
    plt.plot(t_data, y_data, alpha=0.3)
    plt.plot(t_data, y_smooth)
    plt.xlim(max(0, t-5), t+1)
    plt.ylim(-1.5, 1.5)
    plt.title("Real-Time Signal")

    plt.pause(0.01) # 10ms

GitHub Repository

None

References

None