simulate_count_data

Simulate a counts matrix <code>X</code> such that
 <code>X[i,j]</code> is Poisson with rate (mean) <code>Y[i,j]</code>, where
 <code>Y = tcrossprod(L,F)</code>, <code>L</code> is an n x k loadings
 (&#8220;activations&#8221;) matrix, and <code>F</code> is an m x k factors
 (&#8220;basis vectors&#8221;) matrix. The entries of matrix <code>L</code> are
 drawn uniformly at random between zero and <code>lmax</code>, and the
 entries of matrix <code>F</code> are drawn uniformly at random between 0
 and <code>fmax</code>.

Implements fast, scalable optimization algorithms for
fitting topic models ("grade of membership" models) and
non-negative matrix factorizations to count data. The methods
exploit the special relationship between the multinomial topic
model (also, "probabilistic latent semantic indexing") and Poisson
non-negative matrix factorization. The package provides tools to
compare, annotate and visualize model fits, including functions to
efficiently create "structure plots" and identify key features in
topics. The 'fastTopics' package is a successor to the
'CountClust' package. For more information, see
<doi:10.48550/arXiv.2105.13440> and
<doi:10.1186/s13059-023-03067-9>. Please also see the GitHub
repository for additional vignettes not included in the package on
CRAN.

Peter Carbonetto

fastTopics

Fast Algorithms for Fitting Topic Models and Non-Negative Matrix
Factorizations to Count Data

Kevin Luo

Kushal Dey

Joyce Hsiao

Abhishek Sarkar

Anthony Hung

Xihui Lin

Paul C. Boutros

Minzhe Wang

Tracy Ke

Eric Weine

Matthew Stephens

simulate_count_data function

<dl><dt>n</dt>
<dd>Number of rows in simulated count matrix. The number of
rows should be at least 2.</dd>
<dt>m</dt>
<dd>Number of columns in simulated count matrix. The number of
columns should be at least 2.</dd>
<dt>k</dt>
<dd>Number of factors, or &#8220;topics&#8221;, used to determine
Poisson rates. The number of topics should be 1 or more.</dd>
<dt>fmax</dt>
<dd>Factors are drawn uniformly at random between zero and
<code>fmax</code>.</dd>
<dt>lmax</dt>
<dd>Loadings are drawn uniformly at random between zero and
<code>lmax</code>.</dd>
<dt>sparse</dt>
<dd>If <code>sparse = TRUE</code>, convert the counts matrix to
a sparse matrix in compressed, column-oriented format; see
<code><a href="/link/sparseMatrix?package=fastTopics&version=0.6-192" data-mini-rdoc="fastTopics::sparseMatrix">sparseMatrix</a></code>.</dd></dl>

Arguments

Simulate Count Data from Poisson NMF Model — simulate_count_data

<dl>

<dt>n</dt>
<dd>Number of rows in simulated count matrix. The number of
rows should be at least 2.</dd>


<dt>m</dt>
<dd>Number of columns in simulated count matrix. The number of
columns should be at least 2.</dd>


<dt>k</dt>
<dd>Number of factors, or &#8220;topics&#8221;, used to determine
Poisson rates. The number of topics should be 1 or more.</dd>


<dt>fmax</dt>
<dd>Factors are drawn uniformly at random between zero and
<code>fmax</code>.</dd>


<dt>lmax</dt>
<dd>Loadings are drawn uniformly at random between zero and
<code>lmax</code>.</dd>


<dt>sparse</dt>
<dd>If <code>sparse = TRUE</code>, convert the counts matrix to
a sparse matrix in compressed, column-oriented format; see
<code><a href='https://rdrr.io/pkg/Matrix/man/sparseMatrix.html'>sparseMatrix</a></code>.</dd>

</dl>

Simulate Count Data from Poisson NMF Model

simulate_count_data: Simulate Count Data from Poisson NMF Model

Description

Usage

Value

Arguments

Details