dgf: Simulate data generating process

In cjbarrie/extremeR: A Bayesian Hierarchical Case-Control Method for Analyzing Extreme Behavior

Simulate data generating process

Description

Function to simulate data from a relevant data generating process (DGP) ; currently this function supports the creation of DGPs with 1 layer of area effects (i.e. small-area effects)

Usage

dgf(
  n.sims = 1,
  n = 100,
  pi.hat.naive = 0.5,
  p = 1,
  X_corr = 0,
  pi = 0.05,
  Moran.I.corr = 0.8,
  spatial_structure = "scotland_lipcancer"
)

Arguments

n.sims	how many samples of simulated data would you like?;
n	how large (sample size) should each sample be?;
pi.hat.naive	what should be the fraction of cases in the sample?;
p	how many normally-distributed covariates should the DGP have?;
X_corr	what should be the average correlation among these covariates?;
pi	what is the population-level probability of sampling a case?
Moran.I.corr	what degree of global spatial autocorrelation (Moran I) should the underlying DGP have?;
spatial_structure	on which map should the data be simulated ? (scotland_lipcancer, pennsylvania_lungcancer, and newyork_leukemia)

Value

a list object

cjbarrie/extremeR documentation built on April 14, 2023, 9:26 p.m.