randomized_block_krylov_method.hpp

Go to the documentation of this file.

#ifndef MLPACK_METHODS_PCA_DECOMPOSITION_POLICIES_RANDOMIZED_BLOCK_KRYLOV_HPP
#define MLPACK_METHODS_PCA_DECOMPOSITION_POLICIES_RANDOMIZED_BLOCK_KRYLOV_HPP

#include <mlpack/prereqs.hpp>
#include <mlpack/methods/block_krylov_svd/randomized_block_krylov_svd.hpp>

namespace mlpack {
namespace pca {

class RandomizedBlockKrylovSVDPolicy
{
 public:
  RandomizedBlockKrylovSVDPolicy(const size_t maxIterations = 2,
                                 const size_t blockSize = 0) :
      maxIterations(maxIterations),
      blockSize(blockSize)
  {
    /* Nothing to do here */
  }

  void Apply(const arma::mat& data,
             const arma::mat& centeredData,
             arma::mat& transformedData,
             arma::vec& eigVal,
             arma::mat& eigvec,
             const size_t rank)
  {
    // This matrix will store the right singular values; we do not need them.
    arma::mat v;

    // Do singular value decomposition using the randomized block krylov SVD
    // algorithm.
    svd::RandomizedBlockKrylovSVD rsvd(maxIterations, blockSize);
    rsvd.Apply(centeredData, eigvec, eigVal, v, rank);

    // Now we must square the singular values to get the eigenvalues.
    // In addition we must divide by the number of points, because the
    // covariance matrix is X * X' / (N - 1).
    eigVal %= eigVal / (data.n_cols - 1);

    // Project the samples to the principals.
    transformedData = arma::trans(eigvec) * centeredData;
  }

  size_t MaxIterations() const { return maxIterations; }
  size_t& MaxIterations() { return maxIterations; }

  size_t BlockSize() const { return blockSize; }
  size_t& BlockSize() { return blockSize; }

 private:
  size_t maxIterations;

  size_t blockSize;
};

} // namespace pca
} // namespace mlpack

#endif