Interface VcovCLOptions

Multiply each clustering term by G / (G - 1), G its number of clusters. Default true.

The clusters: one label per input row of the fit (the rows the fit dropped for a missing value are dropped here too), an array or a typed array, or for multi-way clustering an array of such arrays, one per clustering dimension. Default: every observation its own cluster.

Set negative eigenvalues of the result to 0. Default false.

Multi-way clustering: use clusters of one observation for the intersection of all dimensions. Default false.

Return the sandwich (default true) or the meat only (false).

"HC0" (alias "HC"), "HC1", "HC2" or "HC3". Default "HC1" for an lm fit and "HC0" for a glm fit, as sandwich.