SOPC

The sparse online principal component can not only process online data sets, but also obtain a sparse solution of online data sets.

Provides a distributed framework for simulating and estimating skew factor models under various skewed and heavy-tailed distributions. The methods support distributed data generation, aggregation of local estimators, and evaluation of estimation performance via mean squared error, relative error, and sparsity measures. The distributed principal component (PC) estimators implemented in the package include 'IPC' (Independent Principal Component),'PPC' (Project Principal Component), 'SPC' (Sparse Principal Component), and other related distributed PC methods. The methodological background follows Guo G. (2023) <doi:10.1007/s00180-022-01270-z>.

Guangbao Guo

DSFM

Distributed Skew Factor Model Estimation Methods

Yu Jin

SOPC function

<dl><dt>data</dt>
<dd>is a highly correlated online data set</dd>
<dt>m</dt>
<dd>is the number of principal component</dd>
<dt>gamma</dt>
<dd>is a sparse parameter</dd>
<dt>eta</dt>
<dd>is the proportion of online data to total data</dd></dl>

Arguments

The sparse online principal component can not only process online data sets, but also obtain a sparse solution of online data sets. — SOPC

<dl>

<dt>data</dt>
<dd>is a highly correlated online data set</dd>


<dt>m</dt>
<dd>is the number of principal component</dd>


<dt>gamma</dt>
<dd>is a sparse parameter</dd>


<dt>eta</dt>
<dd>is the proportion of online data to total data</dd>

</dl>

SOPC: The sparse online principal component can not only process online data sets, but also obtain a sparse solution of online data sets.

Description

Usage

Value

Arguments