【Python, ROS】Real time plotting ROS message with matplotlib

Visualizing data in real time is required often for engineers.

▶日本語のページはコチラ（Japanese ver.）

Outline of the implementation

• Subscribe: ROS message (Float64)
• X axis: time
• Y axis: value of the subscribed message

Source code

#!/usr/bin/env python

import rospy
from std_msgs.msg import Float64
import numpy as np
import matplotlib.pyplot as plt
import time

start_time = 0.0
t_ = 0.0
y_ = 0.0

def callback(msg):
    print('callback')

    global start_time
    global t_
    global y_

    if start_time>0:
        t_ = time.time() - start_time
        y_ = msg.data

def graph():
    print('realtime_graph')
    rospy.init_node('realtime_graph', anonymous=True)
    rospy.Subscriber("/message", Float64, callback)

  #Initialization of the graph
    t = [0 for i in range(100)]	#Size of X axis
    y = [0 for k in range(100)]	#Size of Y axis
    
    plt.ion()
    plt.figure()

    plt.title("y")
    plt.xlabel("time[s]")
    plt.ylabel("y[-]")
    plt.ylim(-10, 10)	#Min of Y axis, max of Y axis (fixed)
    plt.grid(True)
    li, = plt.plot(t, y)
    
    global start_time
    global t_
    global y_

    start_time = time.time()

  #Update of the graph
    while not rospy.is_shutdown():
        print('loop')
        
        t.append(t_)	#Appending a new value in X axis
        t.pop(0)	#Popping the oldest value in X axis
        y.append(y_)	#Appending a new value in Y axis
        y.pop(0)	#Popping the oldest value in Y axis

        li.set_xdata(t)
        li.set_ydata(y)
        plt.xlim(min(t), max(t))	#Min of Y axis, max of Y axis (dynamic)

        plt.draw()
        plt.pause(0.1)	#Update rate
    rospy.spin()

if __name__ == '__main__':
    graph()

Explanation

• Initialization of the graph with zero is require at the beginning.
• When a new message is subscribed, it is appended, and the oldest value is popped.
• Some global variables are needed because callback is special.

Lastly...

A source code which plots subscribed ROS data in real time is proposed in this article. I hope this article will help you, and you can you use it.

Thanks for reading. Good luck!

P.S.

I found that you don’t need to write such a complicate code above after finish writing this article, and you can just use rqt_plot. It is a helpful tool from ROS.