Presets

System preset factories:

• create_presidential_config(...)
• create_parliamentary_config(...)
• create_semi_presidential_config(...)

Use presets for fast comparative experiments across institutional systems.

Why presets are useful

Presets provide:

• faster onboarding,
• consistent institutional assumptions,
• fewer accidental inconsistencies during comparative studies.

They are ideal for baseline scenario design before moving to custom configuration.

Standard comparative pattern

from policyflux import (
    build_engine,
    create_presidential_config,
    create_parliamentary_config,
    create_semi_presidential_config,
)

common = dict(num_actors=120, policy_dim=2, iterations=250, seed=42)

configs = {
    "presidential": create_presidential_config(**common),
    "parliamentary": create_parliamentary_config(**common),
    "semi_presidential": create_semi_presidential_config(**common),
}

for name, config in configs.items():
    engine = build_engine(config)
    engine.run()
    print(f"{name}: {engine.pass_rate:.1%}")

Preset + targeted adjustment

After creating a preset config, apply small focused changes for hypothesis testing.

from policyflux import create_presidential_config

config = create_presidential_config(num_actors=100, policy_dim=2, iterations=200, seed=10)
config = config.with_flat(public_support=0.64)

Keep changes narrow to preserve interpretability.

Preset experiment checklist

1. Keep num_actors, policy_dim, and iterations equal across systems.
2. Use identical seed for direct baseline comparison.
3. Change one additional variable at a time.
4. Record pass-rate deltas and variance across seed sweeps.

For advanced sweeps, continue with Scenarios.