tinit_components: Obtain initial estimates of each mode's components.

Description Usage Arguments Value Author(s)

Description

The inital values of each component are taken to be proportional to the first singular vector of each mode-k matricization of the data array.

Usage

1
2
tinit_components(Y, which_na = NULL, start = c("first_sv", "random"),
  known_factors = NULL, known_modes = NULL)

Arguments

Y

An array of numerics. The data.

which_na

Either NULL (when complete data) or an array of logicals the same dimension as Y, indicating if the observation is missing (TRUE) or observed (FALSE).

start

How should we choose the starting values? Either using the first singular vector along each mode ("first_sv") or randomly ("random").

known_factors

A list of known factors for the modes indicated in known_modes. Defaults to NULL, where all factors are assumed to be unknown.

known_modes

A vector of integers. The modes that are known. Should be the same length as known_factors.

Value

ex_list A list of vectors of the starting expected values of each component.

ex2_vec A vector of starting values of E[x'x]

Author(s)

David Gerard


kkdey/flashr documentation built on May 20, 2019, 10:36 a.m.