leg_count

Sweep a walker design across a range of leg counts.

Post-design leg-count sweeps.

Once a walker design is finalised (topology + link dimensions), the question “how many legs is best?” is typically solved as a small post-hoc sweep rather than as part of the main optimisation loop. sweep_leg_counts() evaluates a fixed design across a range of leg counts and returns the scores, so callers can pick the argmax or plot the trade-off curve.

leggedsnake.leg_count.sweep_leg_counts(walker: Walker, objective: DynamicFitness, n_legs_range: Iterable[int] = range(2, 9), opposite_leg: bool = False, world_config: WorldConfig | None = None) → dict[int, FitnessResult]

Evaluate a finished walker design across a range of leg counts.

For each n in n_legs_range the walker is deep-copied, Walker.add_legs(n - 1)() is called (plus Walker.add_opposite_leg() when requested), and objective is invoked on the resulting topology / dimensions. Results are returned as an ordered mapping so callers can pick the best leg count or plot the trade-off.

The objective is expected to be configured with n_legs=1 (its default add-legs expansion is skipped). If it exposes an n_legs attribute, it is temporarily overridden to 1 for the duration of the call to avoid double-adding legs; the original value is restored before returning.

Parameters:

  • walker (Walker) – Finished single-leg design. Its topology and dimensions are deep-copied per sweep entry; the input walker is not mutated.

  • objective (DynamicFitness) – Fitness evaluator. Built-in objectives like DistanceFitness or EfficiencyFitness all conform to the protocol.

  • n_legs_range (iterable of int) – Leg counts to evaluate. Defaults to range(2, 9).

  • opposite_leg (bool) – If True, Walker.add_opposite_leg() is called before Walker.add_legs(), creating a mirrored-pair baseline (so n_legs=2 means two opposing legs, n_legs=3 means two opposing plus a phase-offset copy, etc.).

  • world_config (WorldConfig, optional) – Simulation config forwarded to the objective.

Returns:

Ordered mapping from leg count to fitness result.

Return type:

dict[int, FitnessResult]