R/rank_transform_effects_matrix.R
rank_transform_effects_matrix.Rd
With compare_cols = FALSE
, transforms each
column of X so that the values lie between 0 and 1, with 0
being the lowest rank, and 1 being the highest. By
“highest rank”, we mean that the entry has the largest
(most positive) value amongst all the entries in the
column. With compare_cols = TRUE
, the “least
extreme” relative ranks are obtained by comparing to the ranks
of the other columns. This results in values between -1 (when
the rank is 0 in the current column and 1 in all other columns)
and +1 (when the rank is 1 in the current column and 0 in all
other columns). When the current column has identical rank to
one or more columns, the “least extreme” relative rank
is zero.
The n x m matrix X.
When compare_cols = TRUE
, instead of
returning the ranks, the “least extreme” relative ranks
are returned.
Describe compare_dims argument here.
The n x m rank-transformed effects matrix.