fastQuantile: linear Algorithm for Quantile Simulation

In linearQ: Linear Algorithm for Simulating Quantiles in Multiscale Change-Point Segmentation Problem

Description

This is a linear algorithm for quantile simulation under null hypothesis in multiscale change-point segmentation.

Usage

fastQuantile(alpha, n, r=round(50/min(alpha, 1-alpha)), 
              mType=c("norm-pen","pois"), seed = 123, ...)	

Arguments

alpha	a scalar with values in [0, 1]; the alpha-quantile of the null distribution of the multiscale statistic via Monte Carlo simulation
n	number of observations
r	number of Monte Carlo simulations
mType	"norm-pen" simulates the multiscale statistic from Normal regression model, "pois" simulates the multiscale statistic from Poission regression model.
seed	data seed
...	further arguments passed to penalty function

Value

A scalar quantile value q.

References

Frick, K., Munk, A., and Sieling, H. (2014). Multiscale Change-Point Inference. J. R. Statist. Soc. B, with discussion and rejoinder by the authors, 76:495–580.

Li, H., Munk, A., and Sieling, H. (2015). FDR-control in multiscale change-point segmentation. arXiv:1412.5844.

Examples

# simulate quantiles for multiscale statistics from Normal regression model
    seed = 123
    q  <- fastQuantile(0.9, 500, 100, mType = "norm-pen")

linearQ documentation built on May 2, 2019, 3:43 a.m.