reconstruction: Full data reconstruction based on posterior samples

In GFA: Group Factor Analysis

Full data reconstruction based on posterior samples

Description

reconstruction returns the full data reconstruction based on given posterior samples.

Usage

reconstruction(res, average = TRUE)

Arguments

res	The sampled model from function gfa
average	If TRUE (default), averages the reconstruction over the posterior predictive samples. If set to FALSE, the output may require a large amount of memory. In case of large input data, we recommend acquiring the posterior predictive samples for subsets of data at a time, based on this implementation.

Value

The data reconstruction, a numeric N \times \sum_{m=1}^M D_m matrix, if average is TRUE (default). Otherwise, the reconstruction is a N \times \sum_{m=1}^M D_m \times Npost array, with posterior samples in the third dimension. If the input data has been paired in two modes, the output will be a list of length 2, one element corresponding to each mode.

GFA documentation built on Oct. 21, 2023, 5:06 p.m.