flash_greedy_init_softImpute: Initialize a flash factor using softImpute

In flashier: Empirical Bayes Matrix Factorization

Initialize a flash factor using softImpute

Description

Initializes a new ("greedy") flash factor using softImpute. When there is missing data, this can yield better results than flash_greedy_init_default without sacrificing much (if any) speed.

Usage

flash_greedy_init_softImpute(flash, seed = 666, ...)

Arguments

flash	A flash_fit object.
seed	Since initialization is random, a default seed is set for reproducibility.
...	Additional parameters to be passed to softImpute.

Value

A list of length two consisting of, respectively, the vector of initial values for loadings \ell_{\cdot k} and the vector of initial factor values f_{\cdot k}.

See Also

flash_greedy, flash_greedy_init_default, flash_greedy_init_irlba

flashier documentation built on Oct. 17, 2023, 5:07 p.m.