add_identifiability

Post-processes the factor scores (H), loadings (B), and intercept (mu)
to ensure a unique solution by applying SVD-based rotation.
This typically enforces orthogonality constraints.

Implements methods for selecting the number of factors in Poisson
factor models, with a primary focus on Thinning Cross-Validation (TCV). The
TCV method is based on the 'data thinning' technique, which probabilistically
partitions each count observation into training and test sets while preserving
the underlying factor structure. The Poisson factor model is then fit on the
training set, and model selection is performed by comparing predictive
performance on the test set. This toolkit is designed for researchers working
with high-dimensional count data in fields such as genomics, text mining, and
social sciences. The data thinning methodology is detailed in Dharamshi et al.
(2025) <doi:10.1080/01621459.2024.2353948> and Wang et al. (2025)
<doi:10.1080/01621459.2025.2546577>.

Zhijing Wang

Determining the Number of Factors in Poisson Factor Models via
Thinning Cross-Validation

Heng Peng

Peirong Xu

add_identifiability function

<dl><dt>H</dt>
<dd>A numeric matrix of factor scores (n x q).</dd>
<dt>B</dt>
<dd>A numeric matrix of factor loadings (p x q).</dd>
<dt>mu</dt>
<dd>A numeric vector for the intercept/mean term.</dd></dl>

Arguments

Enforce Identifiability Constraints on Factor Model Components — add_identifiability

<dl>

<dt>H</dt>
<dd>A numeric matrix of factor scores (n x q).</dd>


<dt>B</dt>
<dd>A numeric matrix of factor loadings (p x q).</dd>


<dt>mu</dt>
<dd>A numeric vector for the intercept/mean term.</dd>

</dl>

add_identifiability: Enforce Identifiability Constraints on Factor Model Components

Description

Usage

Value

Arguments