shrinkage

Shrink a matrix using soft-thresholding or hard-thresholding.

Exploratory principal component analysis for large-scale dataset, including sparse principal component analysis and sparse matrix approximation.

Fan Chen

epca

Exploratory Principal Component Analysis

shrinkage function

<dl><dt>x</dt>
<dd><code>matrix</code> or <code>Matrix</code>, to be threshold.</dd>
<dt>gamma</dt>
<dd><code>numeric</code>, the constraint of Lp norm, i.e. $\|x\|\le \gamma$.</dd>
<dt>shrink</dt>
<dd><code>character(1)</code>, shrinkage method, either "soft"- (default) or "hard"-thresholding (see details).</dd>
<dt>epsilon</dt>
<dd><code>numeric</code>, precision tolerance. This should be greater than <code>.Machine$double.eps</code>.</dd></dl>

Arguments

Shrinkage — shrinkage

<dl>

<dt>x</dt>
<dd><code>matrix</code> or <code>Matrix</code>, to be threshold.</dd>


<dt>gamma</dt>
<dd><code>numeric</code>, the constraint of Lp norm, i.e. $\|x\|\le \gamma$.</dd>


<dt>shrink</dt>
<dd><code>character(1)</code>, shrinkage method, either "soft"- (default) or "hard"-thresholding (see details).</dd>


<dt>epsilon</dt>
<dd><code>numeric</code>, precision tolerance. This should be greater than <code>.Machine$double.eps</code>.</dd>

</dl>

shrinkage: Shrinkage

Description

Usage

Value

Arguments

Details

References

See Also

Examples