DataGen_rare_group {MUGS} | R Documentation |
Function used to generate input data (used only for Simulations) Generate SPPMIs, dummy matrices based on prior group structures, and code-code pairs for tuning and evaluation
Description
Function used to generate input data (used only for Simulations) Generate SPPMIs, dummy matrices based on prior group structures, and code-code pairs for tuning and evaluation
Usage
DataGen_rare_group(
seed = NULL,
p,
n1,
n2,
n.common,
n.group,
sigma.eps.1,
sigma.eps.2,
ratio.delta,
network.k,
rho.beta,
rho.U0,
rho.delta,
sigma.rare,
n.rare,
group.size
)
Arguments
seed |
for reproducibility |
p |
the length of an embedding |
n1 |
the number of codes in site 1 |
n2 |
the number of codes in site 2 |
n.common |
common: the number of overlapping codes |
n.group |
the number of groups |
sigma.eps.1 |
the sd of error in site 1 |
sigma.eps.2 |
the sd of error in site 2 |
ratio.delta |
the proportion of codes in each site that have site-specific effects applied to them |
network.k |
the number of distinct blocks within each site for which unique inter-code correlations are modeled |
rho.beta |
AR parameter for the group effects covariance matrix |
rho.U0 |
AR parameter for the code effects covariance matrix |
rho.delta |
AR parameter for the code-site effects covariance matrix |
sigma.rare |
the sd of error for rare codes (usually larger than sigma.eps.1 and sigma.eps.2) |
n.rare |
The number of rare codes |
group.size |
the size of each group |
Value
Returns input data, SPPMIs, dummy matrices based on prior group structures and code-code pairs for tuning and evaluation