Monte Carlo-simulation
Swedish
Contents
A Monte Carlo-simulation (MC) is similar to a ordinary simulation, but the MC will contain several replicates. The aim of the functionality is to be able to more readily test the effects of stochastic elements in the simulation. The default settings in Heureka is to use deterministic settings, but there are some models with settings that can be changed:
| • | Under the control table Production Model there are some settings that be changed to stochastic behaviour: Ingrowth model, motrality implementation, deterministic simulation. |
| • | The assignment of different control categories within a forest domain and stand selection in priority functions can be stochastic. |
If models are set to be stochastic the outcome will differ between each Monte Carlo-replicate.
To perform a Monte Carlo-simulation you click the Monte Carlo Simulation-button:
The following dialog is similar to the usual TPG-dialog, but will you will also be prompted on how many replicates you want:
The random seed is the random seed that will be used for the MC-simulation. The replicates will always have different seeds. You can read more about these settings at the Heureka wiki.
After simulation the results can be displayed for one replicate at a time, or for the whole MC-simulation. You can find some summary statistics (mean, min, max, standard error and confidence interval) in Results - Summary Reports.
