compute_E.simple: E step

Description

compute_E.simple computes the E step in the simple case where the invert matrix Sigma_YY_inv is given

Usage

compute_E.simple(phylo, times_shared, distances_phylo, process, params_old, masque_data = c(rep(TRUE, attr(params_old, "p_dim") * length(phylo$tip.label)), rep(FALSE, attr(params_old, "p_dim") * phylo$Nnode)), F_moments, Y_data_vec_known, miss = rep(FALSE, attr(params_old, "p_dim") * length(phylo$tip.label)), Y_data, U_tree, ...)

Arguments

phylo

Input tree.

Y_data

: vector indicating the data at the tips

sim

(list) : result of function simulate

Sigma

: variance-covariance matrix, result of function compute_variance_covariance

Sigma_YY_inv

: invert of the variance-covariance matrix of the data

Value

conditional_law_X (list) : list of conditionnal statistics : "expectation" : matrix of size p x (ntaxa+nNodes), with ntaxa fisrt columns set to Y_data (tips), and from ntaxa+1 to conditional expectation of the nodes conditionned to the tips E[Z_j|Y] "variances" : array of size p x p x (ntaxa+nNodes) with ntaxa first matrices of zeros (tips) and conditional variance of the nodes conditionned to the tips Var[Z_j|Y] "covariances" : array of size p x p x (ntaxa+nNodes) with ntaxa first matrices of zeros (tips) and conditional covariance of the nodes and their parents conditionned to the tips Cov[Z_j,Z_pa(j)|Y], with NA for the root. "optimal.values" : matrix of size p x ntaxa+nNodes of optimal values beta(t_j)

Details

This function takes parameters sim, Sigma and Sigma_YY_inv from compute_mean_variance.simple. It uses functions extract.variance_covariance, extract.covariance_parents, and extract_simulate_internal to extract the needed quantities from these objects.