def_initial: Define initial values of parameters for clustering
In USCbiostats/LUCid: Latent Unknown Clustering with Integrated Data

Defines initial values of model parameters in est_lucid, sem_lucid , & tune_lucid fitting.

def_initial(
  init_b = NULL,
  init_m = NULL,
  init_s = NULL,
  init_g = NULL,
  init_pcluster = NULL
)

`init_b`	Initial model parameters of β, genetic effects parameter: `K*(ncol(G)+1)` dimensional matrix, each row refers to a latent cluster and the first column is the intercept.
`init_m`	Initial model parameters of μ, biomarker mean effects parameters: `K*ncol(Z)` dimensional matrix, each row refers to a latent cluster.
`init_s`	Initial model parameters of Σ, biomarker covariance matrix: a list of `K` `ncol(Z)*ncol(Z)` matrices.
`init_g`	Initial model parameters of γ, outcome effects parameter: a vector with a length of `K` for binary Y or `2K` for continuous Y. For binary Y, they are log odds in K clusters; for continuous Y, they are K cluster-specific means followed by standard deviations in K clusters.
`init_pcluster`	Initial probabilities of latent clusters.