Algorithm for generating a steam of generalised pivotal quantities for the regression coefficients. If adjusted=FALSE, then no adjustments are made for the uncertainty in the heteroscedasticity estimates d. If adjusted=TRUE, then adjustments are performed. In this case, 's' needs to be provided.

1 2 | ```
pivotalStream(n, y, d, x, s = NULL,
method = list("univariate", "multivariate"), adjusted)
``` |

`n` |
length of stream. |

`y` |
k-vector of responses. |

`d` |
k-vector of heteroscedasticity. |

`x` |
design (k,p)-matrix. |

`s` |
k-vector of study responses. No need to provide this, when adjusted=FALSE. Default is NULL. |

`method` |
A list. Used to choose the methods for calculating the pivotal quantities of the regression coefficients. Default is 'method=list("univariate", "multivariate")'. |

`adjusted` |
TRUE or FALSE. Default is FALSE. |

If method=="univariate" or method=="multivariate", then the return is a (p+1)-n-matrix. The first row contains pivotal quantities of the heterogeneity, the rest of the rows pivotal quantities of the regression coefficients. Each column is an independent draw.

If 'method==list("univariate", "multivariate")', then the return is a (2p+1)-n-matrix. Of each column, the first element is a pivotal for the heterogeneity, the next 'p' elements is a pivotal vector for the regression coefficients based on "univariate", the last 'p' elements are a pivotal vector for the regression coefficients based on "multivariate"

Questions? Problems? Suggestions? Tweet to @rdrrHQ or email at ian@mutexlabs.com.

All documentation is copyright its authors; we didn't write any of that.