measurefun

Evaluate the smallest cononical correlation (ccor) coefficients or trace statistic between two matrices, where a larger ccor or trace statistic is better.

Generalized factor model is implemented for ultra-high dimensional data with mixed-type variables.
Two algorithms, variational EM and alternate maximization, are designed to implement the generalized factor model,
respectively. The factor matrix and loading matrix together with the number of factors can be well estimated.
This model can be employed in social and behavioral sciences, economy and finance, and genomics,
to extract interpretable nonlinear factors. More details can be referred to
Wei Liu, Huazhen Lin, Shurong Zheng and Jin Liu. (2021) <doi:10.1080/01621459.2021.1999818>.

Wei Liu

Generalized Factor Model

Huazhen Lin

Shurong Zheng

Jin Liu

Jinyu Nie

measurefun function

<dl> <dt>hH</dt>
<dd>a <code>n</code>-by-<code>q</code> matrix, the estimated matrix.</dd> <dt>H</dt>
<dd>a <code>n</code>-by-<code>q</code> matrix, the true matrix.</dd> <dt>type</dt>
<dd>a character taking value within <code>c('trace_statistic','ccor')</code>, default as 'trace_statistic'.</dd></dl>

Arguments

Author

Assess the performance of  an estimator on a matrix — measurefun

<dl>

 <dt>hH</dt>
<dd>a <code>n</code>-by-<code>q</code> matrix, the estimated matrix.</dd>

 <dt>H</dt>
<dd>a <code>n</code>-by-<code>q</code> matrix, the true matrix.</dd>

 <dt>type</dt>
<dd>a character taking value within <code>c('trace_statistic','ccor')</code>, default as 'trace_statistic'.</dd>

</dl>

measurefun: Assess the performance of an estimator on a matrix

Description

Usage

Value

Arguments

Author

Examples