set_as_numeric_matrix {dataPreparation} | R Documentation |
Prepare a numeric matrix from a data.table. This matrix is suitable for
machine learning purposes, since factors are binarized. It may be sparsed,
include an intercept, and drop a reference column for each factor if
required (when using lm()
, for instance)
set_as_numeric_matrix(
data_set,
intercept = FALSE,
all_cols = FALSE,
sparse = FALSE
)
data_set |
data.table |
intercept |
Should a constant column be added? (logical, default to FALSE) |
all_cols |
For each factor, should we create all possible dummies, or should we drop a reference dummy? (logical, default to FALSE) |
sparse |
Should the resulting matrix be of a (sparse) Matrix class? (logical, default to FALSE) |