do_sim: Function to simulate agreement patterns for record pairs from...

In corlink: Record Linkage, Incorporating Imputation for Missing Agreement Patterns, and Modeling Correlation Patterns Between Fields

Description

Function to simulate agreement patterns for record pairs from a mixture model

Usage

do_sim(cor_match = 0.1, cor_mismatch = 0, nsample = 10^3,
  pi_match = 0.01, m_probs, u_probs, missingprobs)

Arguments

cor_match	correlation in 0/1 agreement fields for record pairs constituting the same individual (matches).
cor_mismatch	correlation in 0/1 agreement fields for record pairs not constituting the same individual (true mismatches).
nsample	number of record pairs to simulate.
pi_match	the probability both records from a randomly selected record pair correspond to the same individual (a true match).
m_probs	marginal probabilities of agreement for each field for record pairs constituting true matches.
u_probs	marginal probabilities of agreement for each field for record pairs constituting true mismatches.
missingprobs	probabilities that each field is missing on at least one record pair.

Value

A matrix of the simulated agreement patterns, with final column equal to the count of each pattern.

Examples

m_probs <- rep(0.8,6)
u_probs <- rep(0.2,6)
means_match <- -1*qnorm(1-m_probs)
means_mismatch <- -1*qnorm(1-u_probs)
missingprobs <- rep(.2,6)
thedata <- do_sim(cor_match=0.2,cor_mismatch=0,nsample=10^4,
pi_match=.5,m_probs=rep(0.8,5),u_probs=rep(0.2,5),missingprobs=rep(0.4,5))
thedata

corlink documentation built on May 2, 2019, 12:19 p.m.