Monitoring resource usage

[Parthasarati Dileepan]

documentation at the following link, but it is just too much for me. I am looking for an example of how it is used and how to compute time-based average line length. I appreciate any guidance.

https://simpy.readthedocs.io/en/latest/topical_guides/monitoring.html#monitoring-your-processes

[camable10]

The following lines might help.  Place them somewhere in your code.

Once the simulation ends you can process the list and get the average line length you need.

def monitor(env, data, resource):

  while True:

     item = (

      resource._env.now,  # The current simulation time

      resource.count,  # The number of users

      len(resource.queue),  # The number of queued processes

     )

     data.append(item)

     yield env.timeout(1)   #monitor every 1 time unit

data = []    #list with collected data

env.process(monitor(env, data, resource))   # define the resource to monitor

[rt.van....@gmail.com]

Beware that  the method proposed by @caam...@yahoo.com is just an approximation of the real length of the queue as the length is tallied only at discrete moments. You could improve the accuray by making a smaller step, at the cost of performance and memory, but you will never get the real mean (and other statistics).

If you want to do it properly, you should tally the length at the moment of a change and do the right (relatively complicated) calculations afterward.

The 'other' discrete event simulation package salabim has so called monitors that take care of all that.

[Michael Gibbs]

its not that hard, see my post in stack overflow

Stack overflow monitor example

[is...@oakland.edu]

I use the same approach as michaelgbs3005. By logging tuples of (time, queue length) you can pretty easily do time averaged stats as shown in the StackOverflow monitor example. If you need things like time averaged standard dev or percentiles you can use DescrStatsW from statsmodels.:

from statsmodels.stats.weightstats import DescrStatsW