Fit a multi-group negative-binomial model to SAGE data, with Pearson estimation of the common overdispersion parameter.

1 |

`y` |
numeric response vector. |

`x` |
numeric matrix of covariates, candidates to be added to the regression. |

`xkept` |
numeric matrix of covariates to be included in the starting regression. |

`intercept` |
logical, should an intercept be added to |

`nvar` |
integer, number of covariates from |

This function has the advantage that `x`

can have many more columns than the length of `y`

.

Integer vector of length `nvar`

, giving the order in which columns of `x`

are added to the regression.

Gordon Smyth

`step`

1 2 3 |

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.