Basics of Discrete Event Simulation using SimPy


SimPy is a powerful process-based discrete event simulation framework written in Python.

Installation :
To install SimPy, use the following command –

pip install simpy

Basic Concepts :

The core idea behind SimPy is the generator function in Python. The difference between a normal function and a generator is that a normal function uses the “return” statement, while a generator uses “yield” statement.

If the function has a return statement, then even on multiple function calls, it returns the same value. For eg –



filter_none

edit
close

play_arrow

link
brightness_4
code

def func():
    return 1
    return 2

chevron_right


When the func() is called during the runtime, it will always return at the first instance of the return statement, that is, the function func() always returns 1, and the next return statement is never executed.

However, in discrete event simulation, we may need to find the state of the system at a given time T. For that, it is required to remember the state of the interval just before T, and then perform the given simulation and return to state at time T.

This is where generator functions are quite useful. For example, consider the following function

filter_none

edit
close

play_arrow

link
brightness_4
code

def func():
    while True:
        yield 1
        yield 2

chevron_right


Now, when the first time this function is called, it ‘yields’ 1. However, on the very next call, it will yield 2. In some sense, it remembers what it returned upon the last call, and moves on to the next yield statement.

Events in SimPy are called processes, which are defined by generator functions of their own. These processes take place inside an Environment. (Imagine the environment to be a large box, inside of which the processes are kept.)

Consider a simple example, involving the simulation of a traffic light –

filter_none

edit
close

play_arrow

link
brightness_4
code

# Python 2.7 code to demonstrate basics of SimPy package
# Simulation of a Traffic Light
  
# import the SimPy package
import simpy
  
# Generator function that defines the working of the traffic light
# "timeout()" function makes next yield statement wait for a
# given time passed as the argument
def Traffic_Light(env):
  
    while True:
  
        print "Light turns GRN at " + str(env.now)
          
        # Light is green for 25 seconds
        yield env.timeout(25)        
  
        print "Light turns YEL at " + str(env.now)
          
        # Light is yellow for 5 seconds
        yield env.timeout(5)
  
        print "Light turns RED at " + str(env.now)
          
        # Light is red for 60 seconds
        yield env.timeout(60)
  
# env is the environment variable
env = simpy.Environment()        
  
# The process defined by the function Traffic_Light(env)
# is added to the environment
env.process(Traffic_Light(env))
  
# The process is run for the first 180 seconds (180 is not included)
env.run(until = 180)

chevron_right


Output :

Light turns GRN at 0
Light turns YEL at 25
Light turns RED at 30
Light turns GRN at 90
Light turns YEL at 115
Light turns RED at 120

In this code, the generator function Traffic_Light(env) takes the environment variable as the argument and simulates the operation of the traffic light for the time period passed as argument in the env.run() function. (Actually, time in SimPy is unitless. Though it can be converted to hours, minutes or seconds as per convenience). env.now returns the current value of the time elapsed.

env.timeout() function is the base of this simulation, as it waits for the time passed as the argument to be elapsed on the computer’s simulation clock (it is not a real time clock), and then initiate the next yield statement, till the time passed as argument in env.run() has finished.

env.run() starts all the processes linked to the environment at the same time = 0.




My Personal Notes arrow_drop_up

Check out this Author's contributed articles.

If you like GeeksforGeeks and would like to contribute, you can also write an article using contribute.geeksforgeeks.org or mail your article to contribute@geeksforgeeks.org. See your article appearing on the GeeksforGeeks main page and help other Geeks.

Please Improve this article if you find anything incorrect by clicking on the "Improve Article" button below.


Article Tags :

Be the First to upvote.


Please write to us at contribute@geeksforgeeks.org to report any issue with the above content.