[mlpack-git] master: Code cleanup. (ee69124)

[gitdub at big.cc.gt.atl.ga.us]

Repository : https://github.com/mlpack/mlpack

On branch  : master
Link       : https://github.com/mlpack/mlpack/compare/ab9bc6b3bb911c155d706cb11d505e4cace31915...8b3dd9b9dfabadaeb7d62151162f20e1bf9fbbe3

>---------------------------------------------------------------

commit ee691246a140264051d1f6a53f71257c21e3dcca
Author: Ryan Curtin <ryan at ratml.org>
Date:   Mon Feb 9 13:17:40 2015 -0500

    Code cleanup.


>---------------------------------------------------------------

ee691246a140264051d1f6a53f71257c21e3dcca
 src/mlpack/methods/amf/update_rules/nmf_als.hpp | 56 +++++++++++++++++--------
 1 file changed, 38 insertions(+), 18 deletions(-)

diff --git a/src/mlpack/methods/amf/update_rules/nmf_als.hpp b/src/mlpack/methods/amf/update_rules/nmf_als.hpp
index ea96630..395f93d 100644
--- a/src/mlpack/methods/amf/update_rules/nmf_als.hpp
+++ b/src/mlpack/methods/amf/update_rules/nmf_als.hpp
@@ -14,11 +14,24 @@ namespace amf {
 
 /**
  * This class implements a method titled 'Alternating Least Squares' described
- * in the paper 'Positive Matrix Factorization: A Non-negative Factor Model with 
- * Optimal Utilization of Error Estimates of Data Values' by P Paatero and 
- * U Tapper. It uses least squares projection formula to reduce the error 
- * value of \f$ \sqrt{\sum_i \sum_j(V-WH)^2} \f$ by alternately calculating W 
- * and H respectively while holding the other matrix constant.
+ * in the following paper:
+ *
+ * @code
+ * @article{paatero1994positive,
+ *  title={Positive matrix factorization: A non-negative factor model with
+ *      optimal utilization of error estimates of data values},
+ *  author={Paatero, P. and Tapper, U.},
+ *  journal={Environmetrics},
+ *  volume={5},
+ *  number={2},
+ *  pages={111--126},
+ *  year={1994}
+ * }
+ * @endcode
+ *
+ * It uses the least squares projection formula to reduce the error value of
+ * \f$ \sqrt{\sum_i \sum_j(V-WH)^2} \f$ by alternately calculating W and H
+ * respectively while holding the other matrix constant.
  */
 class NMFALSUpdate
 {
@@ -26,20 +39,25 @@ class NMFALSUpdate
   //! Empty constructor required for the UpdateRule template.
   NMFALSUpdate() { }
 
+  /**
+   * Set initial values for the factorization.  In this case, we don't need to
+   * set anything.
+   */
   template<typename MatType>
-  void Initialize(const MatType& dataset, const size_t rank)
+  void Initialize(const MatType& /* dataset */, const size_t /* rank */)
   {
-      (void)dataset;
-      (void)rank;
+    // Nothing to do.
   }
 
   /**
-   * The update rule for the basis matrix W. The formula used is
+   * The update rule for the basis matrix W. The formula used isa
+   *
    * \f[
-   * W^T = \frac{HV^T}{HH^T}
+   * W^T = \frac{H V^T}{H H^T}
    * \f]
-   * The function takes in all the matrices and only changes the
-   * value of the W matrix.
+   *
+   * The function takes in all the matrices and only changes the value of the W
+   * matrix.
    *
    * @param V Input matrix to be factorized.
    * @param W Basis matrix to be updated.
@@ -54,7 +72,7 @@ class NMFALSUpdate
     // W = (inv(H * H.t()) * H * V.t()).t();
     W = V * H.t() * pinv(H * H.t());
 
-    // Set all negative numbers to machine epsilon
+    // Set all negative numbers to machine epsilon.
     for (size_t i = 0; i < W.n_elem; i++)
     {
       if (W(i) < 0.0)
@@ -66,11 +84,13 @@ class NMFALSUpdate
 
   /**
    * The update rule for the encoding matrix H. The formula used is
+   *
    * \f[
-   * H = \frac{W^TV}{W^TW}
+   * H = \frac{W^T V}{W^T W}
    * \f]
-   * The function takes in all the matrices and only changes the
-   * value of the H matrix.
+   *
+   * The function takes in all the matrices and only changes the value of the H
+   * matrix.
    *
    * @param V Input matrix to be factorized.
    * @param W Basis matrix.
@@ -94,7 +114,7 @@ class NMFALSUpdate
   }
 }; // class NMFALSUpdate
 
-}; // namespace amf
-}; // namespace mlpack
+} // namespace amf
+} // namespace mlpack
 
 #endif