qcov: Quantile Covariance

In abnormally-distributed/cvreg: Cross Validation and Robust Estimation Utilities

Description

This returns the quantile covariance estimator from Li, Li, and Tsai (2015). The Harrell-Davis estimator is used for the quantile estimation. Note that the quantile covariance is not symmetric, hence qcov(x,y) != qcov(y,x). The reason is that this metric does not measure the covariance at the qth quantile for both variables, but rather, the covariance of y with the qth quantile of x.

Usage

qcov(x, y, q = 0.5)

Value

the quantile covariance

References

Harrell, F. & Davis, C. (1982) A new distribution-free quantile estimator. Biometrika, 69, 635-640.

Li, G., Li, Y., & Tsai, C. (2015) Quantile correlations and quantile autoregressive modeling. J Am Stat Assoc. 110:246–261. doi: 10.1080/01621459.2014.892007

abnormally-distributed/cvreg documentation built on May 3, 2020, 3:45 p.m.