Computer Generation of Bivariate Gamma Random Vectors.
Abstract
The use of bivariate gamma distributions in simulation modeling is considered. A family of algorithms is given, any member of which can be used to generate random bivariate vectors having any gamma marginal distributions and any theoretically possible correlation, both positive and negative. Computational considerations are discussed, including the selection of parameters to provide regression curves consistent with the modeler's needs. A modification which is easier to implement and provides most, but not all, correlations is also given. Finally the use of these algorithms to generate first order auoregressive time series is discussed. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Jul 01, 1979
- Accession Number
- ADA072407
Entities
People
- Bruce W. Schmeiser
- Ram Lal
Organizations
- Southern Methodist University