67 – Section on Statistical Computing: Data Science
Evaluation Error Requirements for Generating Random Variates Using Dominated Rejection Algorithms
Timothy Hall
PQI Consulting
This paper provides an implementing analyst with an error analysis framework by which the likelihood and extent of improper, missed, proper, and spurious variate candidates may be generated by a dominated rejection algorithm. If the actual distribution from which the variate values (on a bounded real interval) are being produced significantly differs from the intended distribution (as given by a density function), then whatever statistical inference is attempted for the resulting distribution is invalid. The analytical methods found in this paper provide for objectively quantifying the extent to which spurious and missed variate candidates may either be eliminated, minimized, or tolerated.