Principal Component Analysis with L1 error.
S4 Class implementing PCA with L1 error.
PCA transforms the data so that the L2 reconstruction error is minimized or the variance of the projected data is maximized. This is sensitive to outliers, L1 PCA minimizes the L1 reconstruction error or maximizes the sum of the L1 norm of the projected observations.
funA function that does the embedding and returns a dimRedResult object.
stdparsThe standard parameters for the function.
Dimensionality reduction methods are S4 Classes that either be used
directly, in which case they have to be initialized and a full
list with parameters has to be handed to the @fun()
slot, or the method name be passed to the embed function and
parameters can be given to the ..., in which case
missing parameters will be replaced by the ones in the
@stdpars.
PCA can take the following parameters:
The number of output dimensions.
logical, should the data be centered, defaults to TRUE.
logical, should the data be scaled, defaults to FALSE.
character or function, the method to apply, see the pcaL1 package
other parameters for fun
Wraps around the different methods is the pcaL1 package. Because PCA
can be reduced to a simple rotation, forward and backward projection
functions are supplied.
Park, Y.W., Klabjan, D., 2016. Iteratively Reweighted Least Squares Algorithms for L1-Norm Principal Component Analysis, in: Data Mining (ICDM), 2016 IEEE 16th International Conference On. IEEE, pp. 430-438.
Other dimensionality reduction methods: AutoEncoder-class,
DRR-class,
DiffusionMaps-class,
DrL-class, FastICA-class,
FruchtermanReingold-class,
HLLE-class, Isomap-class,
KamadaKawai-class, LLE-class,
MDS-class, NNMF-class,
PCA-class, UMAP-class,
dimRedMethod-class,
dimRedMethodList, kPCA-class,
nMDS-class, tSNE-class
if(requireNamespace("pcaL1", quietly = TRUE)) {
dat <- loadDataSet("Iris")
## using the S4 Class
pca_l1 <- PCA_L1()
emb <- pca_l1@fun(dat, pca_l1@stdpars)
## using embed()
emb2 <- embed(dat, "PCA_L1")
plot(emb, type = "2vars")
plot(emb@inverse(emb@data), type = "3vars")
}Please choose more modern alternatives, such as Google Chrome or Mozilla Firefox.