Updates a Multiple Correspondence Analysis solution

Share:

Description

This function updates the Multiple Correspondence Analysis (MCA) solution on the indicator matrix using the incremental method of Ross, Lim, Lin, & Yang (2008)

Usage

1
2
## S3 method for class 'i_mca'
update(object, incdata, current_rank, ff = 0, nchunk = 1, ...)

Arguments

object

object of class 'i_mca'

incdata

Matrix of incoming data

current_rank

Rank of approximation; if empty then full rank is used

ff

Number between 0 and 1 indicating the "forgetting factor" used to down-weight the contribution of earlier data blocks to the current solution. When ff = 0 (default) no forgetting occurs

nchunk

Number of data blocks processed up to that point; default is 1

...

Further arguments passed to update

Value

rowpcoord

Row principal coordinates

colpcoord

Column principal coordinates

rowcoord

Row standard coordinates

colcoord

Column standard coordinates

sv

Singular values

inertia.e

Percentages of explained (adjusted) inertia

levelnames

Attribute names

rowctr

Row contributions

colctr

Column contributions

rowcor

Row squared correlations

colcor

Column squared correlations

rowmass

Row masses

colmass

Column masses

indmat

Indicator matrix

m

Number of cases processed up to this point

ff

A copy of ff in the return object

References

Ross, D. A., Lim, J., Lin, R. S., & Yang, M. H. (2008). Incremental learning for robust visual tracking. International Journal of Computer Vision, 77(1-3), 125-141.

Iodice D' Enza, A., & Markos, A. (2015). Low-dimensional tracking of association structures in categorical data, Statistics and Computing, 25(5), 1009-1022.

See Also

add_es, i_mca, plot.i_mca

Examples

1
2
3
4
5
6
7
8
9
data(women, package = "idm")
dat = women[,c(1:4)]
res_mca = i_mca(dat[1:300,])
nchunk = seq(301,2107,258)
for (k in c(1:(length(nchunk)-1)))
{
  res_mca = update(res_mca,dat[c((nchunk[k]+1):nchunk[k+1]),],nchunk=k+1)
}
plot(res_mca, what=c(FALSE, TRUE), animation = FALSE)