benchmark_results

A dataset summarising wall-clock performance for the main PCA entry points
in bigPCAcpp across matrices of increasing size. The benchmarks compare
the classical block-based decomposition, the streaming variant that writes
rotations as it progresses, the scalable stochastic PCA implementation, and
the base R <code><a href="/link/stats%3A%3Aprcomp()?package=bigPCAcpp&version=0.9.0" data-mini-rdoc="bigPCAcpp::stats::prcomp()">stats::prcomp()</a></code> routine for reference.

datasets

High performance principal component analysis routines
that operate directly on 'bigmemory::big.matrix' objects. The
package avoids materialising large matrices in memory by
streaming data through 'BLAS' and 'LAPACK' kernels and provides
helpers to derive scores, loadings, correlations, and
contribution diagnostics, including utilities that stream
results into 'bigmemory'-backed matrices for file-based
workflows. Additional interfaces expose 'scalable' singular value
decomposition, robust PCA, and robust SVD algorithms so that
users can explore large matrices while tempering the influence
of outliers. 'Scalable' principal component analysis is also implemented,
Elgamal, Yabandeh, Aboulnaga, Mustafa, and Hefeeda (2015)
<doi:10.1145/2723372.2751520>.

Frederic Bertrand

bigPCAcpp

Principal Component Analysis for 'bigmemory' Matrices

benchmark_results function

A data frame with 360 rows and 14 columns:<dl>
<dt>dataset</dt>
<dd>Human-readable size label ("small", "medium", "large",
"xlarge").</dd><dt>rows</dt>
<dd>Number of rows in the simulated matrix.</dd><dt>cols</dt>
<dd>Number of columns in the simulated matrix.</dd><dt>ncomp</dt>
<dd>Number of components requested.</dd><dt>method</dt>
<dd>Computation strategy ("classical", "streaming",
"scalable", or "prcomp").</dd><dt>replicate</dt>
<dd>Replication index for repeated runs.</dd><dt>user_time</dt>
<dd>User CPU time in seconds returned by <code><a href="/link/base%3A%3Asystem.time()?package=bigPCAcpp&version=0.9.0" data-mini-rdoc="bigPCAcpp::base::system.time()">base::system.time()</a></code>.</dd><dt>system_time</dt>
<dd>System CPU time in seconds.</dd><dt>elapsed</dt>
<dd>Elapsed (wall-clock) time in seconds.</dd><dt>success</dt>
<dd>Logical flag indicating whether the run completed without
errors.</dd><dt>backend</dt>
<dd>Name of the backend reported by the result object when the
computation succeeded.</dd><dt>iterations</dt>
<dd>Recorded iteration count for iterative methods when
available (otherwise <code>NA</code>).</dd><dt>converged</dt>
<dd>Logical convergence flag for iterative methods when
available.</dd><dt>error</dt>
<dd>Error message captured for failed runs (otherwise <code>NA</code>).</dd>
</dl>

Format

A dataset summarising wall-clock performance for the main PCA entry points
in bigPCAcpp across matrices of increasing size. The benchmarks compare
the classical block-based decomposition, the streaming variant that writes
rotations as it progresses, the scalable stochastic PCA implementation, and
the base R <code><a href='https://rdrr.io/r/stats/prcomp.html'>stats::prcomp()</a></code> routine for reference.

Benchmark timings for bigPCAcpp methods — benchmark_results

A data frame with 360 rows and 14 columns:<dl>
<dt>dataset</dt>
<dd>Human-readable size label ("small", "medium", "large",
"xlarge").</dd>

<dt>rows</dt>
<dd>Number of rows in the simulated matrix.</dd>

<dt>cols</dt>
<dd>Number of columns in the simulated matrix.</dd>

<dt>ncomp</dt>
<dd>Number of components requested.</dd>

<dt>method</dt>
<dd>Computation strategy ("classical", "streaming",
"scalable", or "prcomp").</dd>

<dt>replicate</dt>
<dd>Replication index for repeated runs.</dd>

<dt>user_time</dt>
<dd>User CPU time in seconds returned by <code><a href='https://rdrr.io/r/base/system.time.html'>base::system.time()</a></code>.</dd>

<dt>system_time</dt>
<dd>System CPU time in seconds.</dd>

<dt>elapsed</dt>
<dd>Elapsed (wall-clock) time in seconds.</dd>

<dt>success</dt>
<dd>Logical flag indicating whether the run completed without
errors.</dd>

<dt>backend</dt>
<dd>Name of the backend reported by the result object when the
computation succeeded.</dd>

<dt>iterations</dt>
<dd>Recorded iteration count for iterative methods when
available (otherwise <code>NA</code>).</dd>

<dt>converged</dt>
<dd>Logical convergence flag for iterative methods when
available.</dd>

<dt>error</dt>
<dd>Error message captured for failed runs (otherwise <code>NA</code>).</dd>


</dl>

benchmark_results: Benchmark timings for bigPCAcpp methods

Description

Arguments

Format

Examples