MUGS {MUGS} | R Documentation |
Main function for MUGS algorithm
Description
Main function for MUGS algorithm
Usage
MUGS(
TUNE = FALSE,
Eva = TRUE,
Lambda = c(10),
Lambda.delta = c(1000),
n.core = 4,
tol = 1,
seed = NULL,
S.1 = NULL,
S.2 = NULL,
X.group.source = NULL,
X.group.target = NULL,
pairs.rel.CV = NULL,
pairs.rel.EV = NULL,
p = 100,
n.group = 400,
outdir = NULL
)
Arguments
TUNE |
Logical value indicating whether the function should tune parameters TRUE or use predefined parameters FALSE. |
Eva |
Logical value indicating whether to perform evaluation (TRUE) or skip it (FALSE). |
Lambda |
The candidate values for the tuning parameter controlling the intensity of penalization on the code effects. |
Lambda.delta |
The candidate values for the tuning parameter controlling the intensity of penalization on the code-site effects. |
n.core |
Integer specifying the number of cores to use for parallel processing. |
tol |
Numeric value representing the tolerance level for convergence in the algorithm. |
seed |
Integer used to set the seed for random number generation, ensuring reproducibility. Set to NULL to disable. |
S.1 |
The SPPMI matrix from site 1. |
S.2 |
The SPPMI matrix from site 2. |
X.group.source |
The dummy matrix representing the group structure of codes at site 1. |
X.group.target |
The dummy matrix representing the group structure of codes at site 2. |
pairs.rel.CV |
Code-code pairs used for tuning via cross-validation. |
pairs.rel.EV |
Code-code pairs used for evaluation. |
p |
Integer indicating the length of embeddings. |
n.group |
The number of groups. |
outdir |
Optional directory to write output files. Defaults to a temporary directory. |
Value
A list or saved files containing the embedding matrices, similarity matrices, and site-heterogeneous code analysis.