weighted_mahal: weighted_mahal

In rzgross/uRbanmatching:

Description

Computes weighted Mahalanobis distance, using Choleski decomp.

Usage

weighted_mahal(
  x_mat,
  cov_x,
  weight_vec = NULL,
  treat_vec = NULL,
  sqrt_mahal = TRUE,
  partial_index = NULL
)

Arguments

x_mat	Numeric matrix (adjust non-numeric columns prior), already rank-adjusted if desired.
cov_x	Covariance of x, calculated potentially with ranks.
weight_vec	Vector of weights corresponding to columns of x_mat, giving weights relative to normal Mahalanobis.
treat_vec	optionally specify which units are treated.
sqrt_mahal	Logical, default TRUE; whether to use regular Mahalanobis.
partial_index	How to partition the full matrix.

Value

returns a matrix of pairwise distances; the relevant indexing depends on treat_vec and partial_index

rzgross/uRbanmatching documentation built on Dec. 22, 2021, 8:20 p.m.