| Sumario: | The writer reviews SNP, a computer program for nonparametric time series analysis. SNP is a public-domain package that estimates the conditional density of a stationary multivariate time series, employing a polynomial series expansion and a model selection strategy to calculate the degree of the expansion. In addition, the program includes supplementary features for prediction, plotting, and simulation analysis. The writer concludes that SNP is a highly flexible program for conditional density estimation and deserves to be more widely known.
|
|---|