textmodel_ca implements correspondence analysis scaling on a
dfm. The method is a fast/sparse version of function ca.
textmodel_ca(x, smooth = 0, nd = NA, sparse = FALSE, residual_floor = 0.1)
the dfm on which the model will be fit
a smoothing parameter for word counts; defaults to zero.
Number of dimensions to be included in output; if
retains the sparsity if set to
specifies the threshold for the residual matrix for
calculating the truncated svd.Larger value will reduce memory and time cost
but might reduce accuracy; only applicable when
textmodel_ca() returns a fitted CA textmodel that is a special
class of ca object.
svds in the RSpectra package is applied to enable the fast computation of the SVD.
You may need to set
sparse = TRUE) and
increase the value of
residual_floor to ignore less important
information and hence to reduce the memory cost when you have a very big
If your attempt to fit the model fails due to the matrix being too large,
this is probably because of the memory demands of computing the \(V
\times V\) residual matrix. To avoid this, consider increasing the value of
residual_floor by 0.1, until the model can be fit.
Nenadic, O. & Greenacre, M. (2007). Correspondence Analysis in R, with Two- and Three-dimensional Graphics: The ca package. Journal of Statistical Software, 20(3).
#> Length Class Mode #> sv 7 -none- numeric #> nd 1 -none- numeric #> rownames 14 -none- character #> rowmass 14 -none- numeric #> rowdist 14 -none- numeric #> rowinertia 14 -none- numeric #> rowcoord 98 -none- numeric #> rowsup 0 -none- logical #> colnames 5140 -none- character #> colmass 5140 -none- numeric #> coldist 5140 -none- numeric #> colinertia 5140 -none- numeric #> colcoord 35980 -none- numeric #> colsup 0 -none- logical #> call 2 -none- call