## Seed

Key

**in Model[Simulation] refers to an arbitrary integer that acts as the seed that determines the sequence of generated pseudo-random numbers.**

*Seed*Keeping the same seed when repeating the pricing process ensures the generation of the same sequence, thus leading to the same price result.

Exceptionally the value

**instructs the random generation engine to use clock-based seed every time a new random series is produced.**

*0*This means that setting

**will have the observable effect that different output will be produced every time the simulation runs.**

*0*The downside is that the simulation results will always differ from each other and thus not be reproducible.

If Random Generator =

**, this seed is generally used if the dimensionality of the sample space is bigger than a certain maximum.**

*Low Discrepancy*Then the seed is used to produce random integers that are used to define the directional integers along the extra dimensions, for which no tabulated directional integers exist.

It is also used if Simulation Type =

**and Randomized Runs is greater than 1.**

*DS*Then the seed is used to produce

**random integers, where**

*K-1***is the number of randomized independent sobol simulations specified in Randomized Runs**

*K*These independent sobol simulations are called

**because the numbers in each original sobol sequence are "shifted" by an amount linked to the respective random integer.**

*randomized*The end result is that the

**averages of the**

*K***sobol simulations are related to each other like**

*K***i.i.d. random numbers, which allows us to apply the usual formula for their error estimate.**

*K*