1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
|
#' Relevel factors to a desired level
#'
#' `step_relevel` creates a *specification* of a recipe
#' step that will reorder the provided factor columns so that
#' the level specified by ref_level is first. This is useful
#' for contr.treatment contrasts which take the first level as the
#' reference.
#'
#' @inheritParams step_center
#' @inherit step_center return
#'
#' @param ... One or more selector functions to choose which
#' variables that will be affected by the step. These variables
#' should be character or factor types. See [selections()] for more
#' details.
#' @param role Not used by this step since no new variables are
#' created.
#' @param ref_level A single character value that will be used to
#' relevel the factor column(s) (if the level is present).
#' @param objects A list of objects that contain the information
#' on factor levels that will be determined by [prep.recipe()].
#' @return An updated version of `recipe` with the new step
#' added to the sequence of existing steps (if any).
#'
#' @keywords datagen
#' @concept preprocessing
#' @concept factors
#' @export
#' @details The selected variables are releveled to a level
#' (given by `ref_level`). Placing the `ref_level` in the first
#' position.
#'
#' Note that if the original columns are character, they will be
#' converted to factors by this step.
#'
#'
#' @examples
#'
#' library(modeldata)
#' data(okc)
#' rec <- recipe(~ diet + location, data = okc) %>%
#' step_unknown(diet, new_level = "UNKNOWN") %>%
#' step_relevel(diet, ref_level = "UNKNOWN") %>%
#' prep()
#'
#' data <- bake(rec, okc)
#' levels(data$diet)
step_relevel <-
function(recipe,
...,
role = NA,
trained = FALSE,
ref_level,
objects = NULL,
skip = FALSE,
id = rand_id("relevel")) {
add_step(
recipe,
step_relevel_new(
terms = ellipse_check(...),
role = role,
trained = trained,
ref_level = ref_level,
objects = objects,
skip = skip,
id = id
)
)
}
step_relevel_new <-
function(terms, role, trained, ref_level, objects, skip, id) {
step(
subclass = "relevel",
terms = terms,
role = role,
trained = trained,
ref_level = ref_level,
objects = objects,
skip = skip,
id = id
)
}
#' @export
prep.step_relevel <- function(x, training, info = NULL, ...) {
col_names <- eval_select_recipes(x$terms, training, info)
col_check <- dplyr::filter(info, .data$variable %in% col_names)
if (any(col_check$type != "nominal")) {
rlang::abort(
"Columns must be character or factor: ",
paste0(col_check$variable[col_check$type != "nominal"],
collapse = ", "
)
)
}
# Get existing levels and their factor type (i.e. ordered)
objects <- lapply(training[, col_names], get_existing_values)
# Check to make sure that no ordered levels are provided
order_check <- map_lgl(objects, attr, "is_ordered")
if (any(order_check)) {
rlang::abort(
"Columns contain ordered factors (which cannot be releveled) '",
x$ref_level, "': ",
paste0(names(order_check)[order_check], collapse = ", ")
)
}
# Check to make sure that the reference level exists in the factor
ref_check <- map_lgl(objects, function(x, y) !y %in% x,
y = x$ref_level
)
if (any(ref_check)) {
rlang::abort(
"Columns must contain the reference level '",
x$ref_level, "': ",
paste0(names(ref_check)[ref_check], collapse = ", ")
)
}
step_relevel_new(
terms = x$terms,
role = x$role,
trained = TRUE,
ref_level = x$ref_level,
objects = objects,
skip = x$skip,
id = x$id
)
}
#' @export
bake.step_relevel <- function(object, new_data, ...) {
for (i in names(object$objects)) {
new_data[[i]] <- stats::relevel(as.factor(new_data[[i]]), ref = object$ref_level)
}
if (!is_tibble(new_data)) {
new_data <- as_tibble(new_data)
}
new_data
}
print.step_relevel <-
function(x, width = max(20, options()$width - 30), ...) {
cat("Re-order factor level to ref_level for ", sep = "")
printer(names(x$objects), x$terms, x$trained, width = width)
invisible(x)
}
#' @rdname step_relevel
#' @param x A `step_relevel` object.
#' @export
tidy.step_relevel <- function(x, ...) {
if (is_trained(x)) {
res <- tibble(
terms = names(x$objects),
value = x$ref_level
)
} else {
term_names <- sel2char(x$terms)
res <- tibble(
terms = term_names,
value = x$ref_level
)
}
res$id <- x$id
res
}
|
