ordr (version 0.1.2)

tbl_ord: A unified ordination object class

Description

These functions wrap ordination objects in the class tbl_ord, create tbl_ords directly from matrices, and test for the class and basic structure.

Usage

as_tbl_ord(x)
# S3 method for tbl_ord
as_tbl_ord(x)
make_tbl_ord(rows = NULL, cols = NULL, ...)
is_tbl_ord(x)
is.tbl_ord(x)
valid_tbl_ord(x)
un_tbl_ord(x)

Value

A tbl_ord (as*(), make*()), an S3-class model object that can be wrapped as one (un*()), or a logical value (is*(), value*()).

Arguments

x

An ordination object.

rows, cols

Matrices to be used as factors of a tbl_ord.

...

Additional elements of a custom tbl_ord.

Details

The tbl_ord class wraps around a range of ordination classes, making available a suite of ordination tools that specialize to each original object class. These tools include format() and ggplot2::fortify() methods, which facilitate the print() method and the ggbiplot() function.

No default method is provided for as_tbl_ord(), despite most defined methods being equivalent (simply appending 'tbl_ord' to the vector of object classes). This prevents objects for which other methods are not defined from being re-classed as tbl_ords.

The function make_tbl_ord() creates a tbl_ord structured as a list of two matrices, u and v, which must have the same number of columns and the same column names.

is_tbl_ord() checks an object x for the tbl_ord class; valid_tbl_ord() additionally checks for consistency between recover_coord(x) and the columns of recover_rows(x) and recover_cols(x), using the recoverers. un_tbl_ord() removes attributes associated with the tbl_ord class in order to restore an object that was originally passed to as_tbl_ord.

Examples

# illustrative ordination: FA of Swiss social data
swiss_fa <- factanal(swiss, factors = 3L, scores = "regression")
print(swiss_fa)

# add the 'tbl_ord' wrapper
swiss_fa_ord <- as_tbl_ord(swiss_fa)
# inspect wrapped model
is_tbl_ord(swiss_fa_ord)
print(swiss_fa_ord)
valid_tbl_ord(swiss_fa_ord)
# unwrap the model
un_tbl_ord(swiss_fa_ord)

# create a 'tbl_ord' directly from row and column factors
# (missing inertia & other attributes)
swiss_fa_ord2 <- make_tbl_ord(rows = swiss_fa$scores, cols = swiss_fa$loadings)
# inspect wrapped factors
is_tbl_ord(swiss_fa_ord2)
print(swiss_fa_ord2)
valid_tbl_ord(swiss_fa_ord2)
# unwrap factors
un_tbl_ord(swiss_fa_ord2)