OptStiefelGBB: A feasible method for optimization with orthogonality...

In kusakehan/CLEMM: Package for paper CLEMM:Clustering and Envelope Mixture Models

Description

Curvilinear search algorithm for optimization on Stiefel manifold

Usage

OptManiMulitBallGBB(X, opts, fun, ...)

Arguments

X	n by k matrix such that X'*X = I
fun	objective function and its gradient: # [F, G] = fun(X, data1, data2) # F, G are the objective function value and gradient, repectively # data1, data2 are addtional data, and can be more
opts	option structure with fields: # record = 0, no print out # mxitr max number of iterations # xtol stop control for ||X_k - X_k-1|| # gtol stop control for the projected gradient # ftol stop control for |F_k - F_k-1|/(1+|F_k-1|) # usually, maxxtol, gtol > ftol

Details

Used in clemm_em function

Value

X	solution
Out	output information

References

Wen, Z., & Yin, W. (2013). A feasible method for optimization with orthogonality constraints. Mathematical Programming, 142(1-2), 397-434.

Examples

Gammaest <- OptStiefelGBB(Gammaest_init, opts, FGfun, p, Sx, S_tmp)$X

kusakehan/CLEMM documentation built on May 24, 2019, 2:46 p.m.