rsci_model: Create a regime switching cointegrated vector error...
In EmilNejst/rscivem:

Create a regime switching cointegrated vector error correction

rsci_model(dim, rank, lags, nreg, pars, fn_prob, beta = NULL,
  lambda = NULL, Phi = NULL, Omega = NULL,
  linres_Phi = rsci_build_linres_Phi(dim, rank, lags, nreg),
  linres_Omega = rsci_build_linres_Omega(dim, nreg))

`dim`	integer, the number of dimentions in the data
`rank`	integer, the cointegration rank.
`lags`	integer, the number of lags in levels in the model.
`nreg`	integer, the number of regimes.
`pars`	vector, parameters that enter the probability function to generate regime probabilities.
`fn_prob`	function, a function that takes at least two arguments `fn_prob(pars, data_struct)` where pars is a the vector of cointegration and probability parameters. data_struct is a rsci_data object.
`beta`	matrix, the matrix of cointegration relations
`lambda`	vector, the vector of probability parameters
`Phi`	list, a list with Phi matrices for all regimes. Default is NULL in which case a standardized procedure is used for estimating initial values.
`Omega`	list, a list with the Omega matrices for all regimes. Default is NULL in which case a standardized procedure is used.
`linres_Phi`	list, a list of an H matrix and an h vector for imposing linear restrictions on a vector given by, vec(Φ) = vec((Φ_1, Φ_2,..., Φ_m)) such that vec(Φ) = Hρ + h where m refers to the number of regimes and ρ . This use of linear restrictions allows the user to impose cross regime restrictions. That is useful if one wishes certain elements to be fixed across regimes. NULL is default resulting in an unrestricted system.
`linres_Omega`	list, a list of an H matrix and an h vector for imposing linear restriction on a vector given by, vec(Ω) = (vec(Ω_1), vec(Ω_2),...,vec(Ω_m)) such that vec(Ω) = Hν + h where ν is the vector of freely varying parameters in Omega. NULL is default and will result in an unrestricted system.