Dynamic Monte Carlo methods for the SAW

In any case, the incomplete-enumeration algorithm is potentially quite efficient, and deserves a systematic theoretical and empirical study. 

In this section I discuss dynamic Monte Carlo methods for generating SAWs in various ensembles. I emphasize that the stochastic dynamics is merely a numerical algorithm, whose goal is to provide statistically independent samples from the desired distribution tt in the smallest CPU time possible. It need not correspond to any real physical dynamics Indeed, the most efficient algorithms typically make non-local moves which would be impossible for real polymer molecules. 

This subsection contains some exceedingly pedantic—but I hope useful— general considerations on dynamic Monte Carlo algorithms. 

The study of a Monte Carlo algorithm can be divided into three stages  

To specify a dynamic Monte Carlo algorithm, we must specify three things  

---

For the most part, the statistical analysis of SAW Monte Carlo data uses the same methods as are employed in other types of Monte Carlo simulation. In particular, with dynamic Monte Carlo data it is essential to carry out a thorough autocorrelation analysis, only in this way can one test the adequacy of the thermalization interval and the run length and properly assess the statistical error bars. For details, see e.g.. Ref. 36, Section 3 Ref. 37, Section 2 Ref. 96, Appendix C and Ref. 11, Sections 9.2.2 and 9.2.3. 

---