LSSPCA-package: LSSPCA: A Function to compute Least Squares Sparse Principal...

In merolagio/LSSPCA: Computes Least Squares Sparse Principal Components

Description

This package is an appendix to the paper my paper and comes with only function. Sparse principal components are combinations of only few of the observed variables. The LS SPCA components give the best possible data approximation to the data under sparsity constraints. The lsspca function takes a data matrix and a goodness of fit specification, (percent variance explained) or a cardinality. Optimal orthogonal components are computed choosing method = "u", suboptimal correlated components with method = "c" and less computationally demanding ones with method = "p".

Details

Subsets of variables can be chosen with different search algorithms and variables can be foced in or out from these subsets.

References

Giovanni M. Merola. 2014. Least Squares Sparse Principal Component Analysis: a Backward Elimination approach to attain large loadings. Austr.&NZ Jou. Stats. 57, pp 391-429

Giovanni M. Merola and Gemai Chen. 2019. Sparse Principal Component Analysis: an efficient Least Squares approach. Jou. Multiv. Analysis 173, pp 366–382 http://arxiv.org/abs/1406.1381

See Also

lsspca for usage examples.

merolagio/LSSPCA documentation built on April 29, 2021, 4:17 p.m.