How can I dynamically update a single matplotlib plot like MATLAB's drawnow?

Below is one way to replicate the MATLAB dynamic update in Python using matplotlib’s interactive mode. You can create a figure with pre-defined line objects and update their data in each iteration. For example:

PYTHONCopied
import numpy as np
import matplotlib.pyplot as plt

# Example intlag function; replace with your actual computation.
def intlag(n):
    # Dummy function to simulate MATLAB function 
    # Returns: error, t, original values f, x values, interpolant values p.
    x = np.linspace(-1, 1, 200)
    f = np.sin(x)
    # For demonstration, let's vary the interpolant with the iteration count.
    p = np.sin(x) + 0.1 * n * np.cos(5*x)
    err = 0  # Dummy error
    t = n
    return err, t, f, x, p

# Turn interactive mode on.
plt.ion()
fig, ax = plt.subplots()

# Initialize two line objects for the original function and the interpolant.
line_f, = ax.plot([], [], '-r', label='fonction')
line_p, = ax.plot([], [], '-b', label='interpolant')

# Set static axis properties and legend.
ax.axis([-1, 1, -2, 2])
ax.legend()

# Run the loop to update the plot dynamically.
for n in range(1, 101):
    err, t_val, f, x, p = intlag(n)
    
    # Update the data of each line object.
    line_f.set_data(x, f)
    line_p.set_data(x, p)
    
    # Update the title.
    ax.set_title(f't = {n}')
    
    # Redraw the canvas.
    fig.canvas.draw()
    fig.canvas.flush_events()
    
    # Pause to control the update rate.
    plt.pause(0.1)

# Optionally, turn interactive mode off and show the final static plot.
plt.ioff()
plt.show()

Explanation

• Interactive Mode:
  plt.ion() turns on interactive mode so that changes update the existing figure without blocking the script.

• Predefined Lines:
  Instead of re-creating the plot each time, create line objects (line_f and line_p) once. Then in each loop iteration, update their data using set_data().

• Redraw & Pause:
  fig.canvas.draw() and fig.canvas.flush_events() force the figure to update immediately after the data changes. plt.pause(0.1) pauses briefly to allow the GUI event loop to process and gives the appearance of a smooth animation.

• Final Display:
  After the loop ends, turning off interactive mode with plt.ioff() and calling plt.show() will prevent the script from exiting and allow you to inspect the final state.

Replace the dummy intlag function with your actual function to perform the interpolation and generate the required data.