<< kcmeans (X) One matrix of observations/variables pcaapply >>

fact >> (X) One matrix of observations/variables > outlier

outlier

calculation of 3 parameters: T2-Hotelling, Q or residual variances, and leverage, useful to identify outliers

Calling sequence

[t2,q,leverage] = outlier(x,x_scores,x_loadings,lv)

Arguments

x:

x is a Div structure or a matrix of dimensions (n x q)

x_scores:

the scores obtained after a PCA onto x

x_scores is a Div structure or a matrix of dimensions (n x naxes)

x_loadings:

the loadings obtained after a PCA onto x

x_loadings is a Div structure or a matrix of dimensions (n x naxes)

lv:

the maximum number of eigenvectors or latent variables to be computed

lv must be lower or equal to naxes

t2:

Hotelling's T2; a Div structure

T2.d is of dimensions (n x lv)

q:

the residual variances; a Div structure

q.d is of dimensions (n x lv)

leverage:

the leverage effect; a Div structure

leverage.d is of dimensions (n x lv)

Examples

[t2hot,q_var,lever]=outlier(x,x_scores,x_eigenvec,20)

Authors


Report an issue
<< kcmeans (X) One matrix of observations/variables pcaapply >>