Commit | Line | Data |
---|---|---|
086ca318 BA |
1 | #' computeGridLambda |
2 | #' | |
d1531659 | 3 | #' Construct the data-driven grid for the regularization parameters used for the Lasso estimator |
086ca318 | 4 | #' |
d1531659 | 5 | #' @param phiInit value for phi |
e3f2fe8a | 6 | #' @param rhoInit value for rho |
e166ed4e | 7 | #' @param piInit value for pi |
d1531659 | 8 | #' @param gamInit value for gamma |
e3f2fe8a | 9 | #' @param X matrix of covariates (of size n*p) |
10 | #' @param Y matrix of responses (of size n*m) | |
11 | #' @param gamma power of weights in the penalty | |
086ca318 BA |
12 | #' @param mini minimum number of iterations in EM algorithm |
13 | #' @param maxi maximum number of iterations in EM algorithm | |
14 | #' @param tau threshold to stop EM algorithm | |
15 | #' | |
d1531659 | 16 | #' @return the grid of regularization parameters |
086ca318 | 17 | #' |
d1531659 | 18 | #' @export |
0eb161e3 | 19 | computeGridLambda = function(phiInit, rhoInit, piInit, gamInit, X, Y, |
aa480ac1 | 20 | gamma, mini, maxi, tau, fast=TRUE) |
39046da6 | 21 | { |
e166ed4e BA |
22 | n = nrow(X) |
23 | p = dim(phiInit)[1] | |
24 | m = dim(phiInit)[2] | |
25 | k = dim(phiInit)[3] | |
26 | ||
19041906 | 27 | list_EMG = EMGLLF(phiInit, rhoInit, piInit, gamInit, mini, maxi, |
aa480ac1 | 28 | gamma, lambda=0, X, Y, tau, fast) |
e166ed4e BA |
29 | grid = array(0, dim=c(p,m,k)) |
30 | for (i in 1:p) | |
31 | { | |
32 | for (j in 1:m) | |
33 | grid[i,j,] = abs(list_EMG$S[i,j,]) / (n*list_EMG$pi^gamma) | |
34 | } | |
35 | grid = unique(grid) | |
086ca318 | 36 | grid |
39046da6 | 37 | } |