Random Walk |
||
| Let
us recapitulate our findings so far. From the random numbers we obtained a configuration,
here a random walk. This configuration was generated independently from previous
configurations to make a sample of configurations. We call this a sample generated by the simple
sampling method. At this point it is clear why we need uncorrelated random numbers. We need the absents from correlation to generate a true random walk. Every correlation in the random numbers must appear as a bias in the walks. This bias carries over to the observables computed form a simulation. |
||
 |
 |
|