NMFSeed-class


  This class implements a simple wrapper strategy object
  that defines a unified interface to seeding methods, that
  are used to initialise NMF models before fitting them
  with any NMF algorithm.


Provides a framework to perform Non-negative Matrix
Factorization (NMF). The package implements a set of already published algorithms
and seeding methods, and provides a framework to test, develop and plug
new/custom algorithms. Most of the built-in algorithms have been optimized
in C++, and the main interface function provides an easy way of performing
parallel computations on multicore machines.

Renaud Gaujoux

Algorithms and Framework for Nonnegative Matrix Factorization
(NMF)

NMFSeed-class function

<dl> <dt>name</dt><dd>character string giving the name of the
  seeding strategy</dd> <dt>method</dt><dd>workhorse function that implements the
  seeding strategy. It must have signature
  <code>(object="NMF", x="matrix", ...)</code> and initialise the
  NMF model <code>object</code> with suitable values for fitting
  the target matrix <code>x</code>.</dd> </dl> 

Slots

<dl> <dt>algorithm</dt><dd><code>signature(object = "NMFSeed")</code>:
  Returns the workhorse function of the seeding method
  described by <code>object</code>. </dd> <dt>algorithm&lt;-</dt><dd><code>signature(object = "NMFSeed",
  value = "function")</code>: Sets the workhorse function of the
  seeding method described by <code>object</code>. </dd> <dt>NMFSeed</dt><dd><code>signature(key = "NMFSeed")</code>: Creates
  an <code>NMFSeed</code> based on a template object
  (Constructor-Copy), in particular it uses the
  <strong>same</strong> name. </dd> <dt>show</dt><dd><code>signature(object = "NMFSeed")</code>: Show
  method for objects of class <code>NMFSeed</code> </dd> </dl> 

NMFSeed-class: Base class that defines the interface for NMF seeding methods.

Description

Arguments

Slots

Methods