Skip to contents

Computes the maximum of standardized statistics among each stratum across multiple methods.

Usage

max_comb_matrix_block_stratum(
  Z,
  Y,
  block,
  c,
  methods.list.all,
  scores.list.all = NULL,
  block.sum = NULL,
  weight
)

Arguments

Z

An n-dimensional binary treatment assignment vector.

Y

An n-dimensional observed outcome vector.

block

An n-dimensional vector specifying block membership.

c

Threshold for the null hypothesis.

methods.list.all

A list of H method specifications.

scores.list.all

Optional pre-computed scores.

block.sum

Optional pre-computed block summary.

weight

A B-dimensional vector of block weights.

Value

A list of B matrices, one per block, containing the maximum standardized test statistic values.