If the Analysis Type = Probabilistic in the Project Settings dialog, then the following options will be enabled in Project Settings:
Sampling Method
Number of Samples
Pseudo-Random Sampling
Sampling Method
The Sampling Method is the algorithm the program uses to generate samples for each statistical simulation. The program can use either Monte Carlo or Latin Hypercube methods. The Latin Hypercube technique produces comparable results to the Monte Carlo technique, but with fewer samples. Typically, an analysis with 1000 samples generated by the Latin Hypercube method will produce comparable results to an analysis using 5000 samples produced by the Monte Carlo method.
For practical reasons the maximum Number of Samples is limited to 50,000.
Whenever a user makes changes to tunnel, rock or support parameters, or clicks the Compute button, RocSupport generates input distributions using random numbers, and runs a full analysis of the model. If the Pseudo-Random Sampling option is selected in Project Settings, RocSupport uses the same seed to generate random numbers each time. As a result, the same sequence of random numbers is generated with every run. This option is useful if the user needs to reproduce simulation results.
When a user de-selects or turns off the Pseudo-Random Sampling option, RocSupport uses system time to initialize its random number generator. Consequently, every simulation run creates a unique set of random numbers, and produces distinctive results.