initial
The initial reorder point r applied at the start of each replication.
Controls configure replication INITIAL conditions: the current policy parameters are re-seeded from the initial values at each beforeReplication(), so every replication begins under the same settings. Changing an initial value during a replication is therefore an error (guarded); the current parameters remain changeable within a replication via setPolicyParameters (e.g. by future dynamic policies).
The order-up-to level is parameterized as S = r + SDelta (see initialOrderUpToPointDelta), following the RDelta precedent of the (r,Q) policy: every clamped (r, SDelta >= 1) combination satisfies r < S by construction, so no cross-field validation is needed at control-set time and optimizers see a box-constrained space. S >= 1 is still validated when the initial parameters are applied at replication start.