Confidence Interval for Metrics and Simulation

This section describes how to generate a confidence interval for simulation metrics using the run_confidence_interval method. This method performs multiple replications of a simulation, computes relevant metrics for each replication, and then derives the corresponding confidence interval.

• confidence_level (float, optional): Specifies the desired confidence level (for instance, 0.95 for 95%). The default value is 0.95.
• number_of_simulations (int, optional): Indicates how many simulations are executed in each replication. The default value is 1.
• replications (int, optional): Determines how many times the entire simulation process is replicated. The default value is 30.

Below is an example of how to call the method:

# Generating a confidence interval for simulation metrics
simulation.run_confidence_interval(
    confidence_level=0.99,
    number_of_simulations=100,
    replications=40,
)

The resulting table provides several columns for each metric, including a lower boundary (LB), an average (AVG), and an upper boundary (UB). For example:

Metrics	LB - Value	AVG - Value	UB - Value
Avg. Total Simulation Time	78.08	78.52	79.04
...	...	...	...

In this context:

• LB refers to the lower boundary of the confidence interval.
• AVG refers to the mean value calculated across all replications.
• UB refers to the upper boundary of the confidence interval.