Simulation Settings¶

Configuration for simulation timing and parameters.

Schema¶

simulation:
  duration_minutes: <float>        # Required
  warmup_minutes: <float>          # Default: 0
  time_unit: <string>              # Default: "seconds"

Fields¶

duration_minutes¶

Type: float Required: Yes

Total simulation duration in minutes.

simulation:
  duration_minutes: 60  # 1-hour simulation

Guidelines:

Short (1-10 min): Quick tests, debugging
Medium (30-60 min): Standard experiments
Long (120+ min): Statistical stability, rare events

warmup_minutes¶

Type: float Default: 0

Warmup period before metrics collection begins.

simulation:
  duration_minutes: 60
  warmup_minutes: 10  # 10 minutes warmup, 50 minutes measured

Purpose:

Allows system to reach steady state
Eliminates startup transients from metrics
Provides realistic initial conditions

Guidelines:

Scenario	Recommended Warmup
Quick tests	0-1 min
Standard	5-10 min
Baseline comparisons	10-20 min
Long-duration	10-20% of duration

time_unit¶

Type: string Default: "seconds" Options: "seconds", "minutes"

Internal time unit for the simulation.

simulation:
  duration_minutes: 60
  time_unit: seconds  # Default

Note

This is primarily internal. Most user-facing values are already in appropriate units.

Examples¶

Quick Test¶

simulation:
  duration_minutes: 5
  warmup_minutes: 1

Standard Experiment¶

simulation:
  duration_minutes: 60
  warmup_minutes: 10

Reproducible Baseline¶

simulation:
  duration_minutes: 120
  warmup_minutes: 60

Long-Running Analysis¶

simulation:
  duration_minutes: 480  # 8 hours
  warmup_minutes: 60

Seed¶

The random seed is specified at the top level, not in the simulation section:

seed: 42  # Random seed for reproducibility

simulation:
  duration_minutes: 60
  warmup_minutes: 10

Seed behavior:

Same seed + same config = identical results
Different seeds = different random sequences
Override at runtime with --seed flag

Effective Duration¶

Metrics are only collected during the effective measurement period:

Effective duration = duration_minutes - warmup_minutes

Example:

simulation:
  duration_minutes: 60
  warmup_minutes: 10
# Effective measurement: 50 minutes

Performance Considerations¶

Duration Impact¶

Duration	Typical Use	Run Time (Release Build)
1 min	Unit tests	< 1 second
5 min	Quick iteration	1-5 seconds
30 min	Standard test	5-30 seconds
60 min	Full simulation	10-60 seconds
480 min	Long analysis	1-5 minutes

Actual times depend on scenario complexity and hardware.

Statistical Stability¶

Longer durations provide more stable statistics:

Short runs: High variance, quick iteration
Long runs: Stable means, reliable for comparison

For statistical comparisons, ensure adequate duration or use multiple replications.