crosspart_GLS {remotePARTS}

Calculate cross-partition statistics in a partitioned GLS

Description

Calculate cross-partition statistics between two GLS partitions

Usage

crosspart_GLS(
  xxi,
  xxj,
  xxi0,
  xxj0,
  invChol_i,
  invChol_j,
  Vsub,
  nug_i,
  nug_j,
  df1,
  df2,
  small = TRUE,
  ncores = NA
)

Arguments

xxi

numeric matrix xx from partition i

xxj

numeric matrix xx from partition j

xxi0

numeric matrix xx0 from partition i

xxj0

numeric matrix xx0 from partition j

invChol_i

numeric matrix invcholV from partition i

invChol_j

numeric matrix invcholV from partition j

Vsub

numeric variance matrix for Xij (upper block)

nug_i

nugget from partition i

nug_j

nugget from partition j

df1

first degree of freedom

df2

second degree of freedom

small

logical: if TRUE, only return rcoefij, rSSRij, and rSSEij

ncores

an optional integer indicating how many CPU threads to use for matrix calculations.

Value

crosspart_GLS returns a list of cross-partition statistics.

If small = FALSE, the list contains the following elements

Rij

Hi

Hj

Hi0

Hj0

SiR

SjR

rcoefij

rSSRij

rSSEij

Vcoefij

If small = FALSE, the list only contains the necessary elements rcoefij, rSSRij, and rSSEij.

See Also

Other partitionedGLS: MC_GLSpart(), sample_partitions()