-
Notifications
You must be signed in to change notification settings - Fork 107
/
num2factor.R
215 lines (198 loc) · 5.39 KB
/
num2factor.R
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
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
#' Convert Numbers to Factors
#'
#' `step_num2factor` will convert one or more numeric vectors to factors
#' (ordered or unordered). This can be useful when categories are encoded as
#' integers.
#'
#' @inheritParams step_center
#' @param transform A function taking a single argument `x` that can be used
#' to modify the numeric values prior to determining the levels (perhaps using
#' [base::as.integer()]). The output of a function should be an integer that
#' corresponds to the value of `levels` that should be assigned. If not an
#' integer, the value will be converted to an integer during [bake()].
#' @param levels A character vector of values that will be used as the levels.
#' These are the numeric data converted to character and ordered. This is
#' modified once [prep()] is executed.
#' @param ordered A single logical value; should the factor(s) be ordered?
#' @template step-return
#' @details
#'
#' # Tidying
#'
#' When you [`tidy()`][tidy.recipe()] this step, a tibble with columns
#' `terms` (the selectors or variables selected) and `ordered` is returned.
#'
#' @template case-weights-not-supported
#'
#' @family dummy variable and encoding steps
#' @export
#' @examplesIf rlang::is_installed("modeldata")
#' library(dplyr)
#' data(attrition, package = "modeldata")
#'
#' attrition %>%
#' group_by(StockOptionLevel) %>%
#' count()
#'
#' amnt <- c("nothin", "meh", "some", "copious")
#'
#' rec <-
#' recipe(Attrition ~ StockOptionLevel, data = attrition) %>%
#' step_num2factor(
#' StockOptionLevel,
#' transform = function(x) x + 1,
#' levels = amnt
#' )
#'
#' encoded <- rec %>%
#' prep() %>%
#' bake(new_data = NULL)
#'
#' table(encoded$StockOptionLevel, attrition$StockOptionLevel)
#'
#'
#' # an example for binning
#'
#' binner <- function(x) {
#' x <- cut(x, breaks = 1000 * c(0, 5, 10, 20), include.lowest = TRUE)
#' # now return the group number
#' as.numeric(x)
#' }
#'
#' inc <- c("low", "med", "high")
#'
#' rec <-
#' recipe(Attrition ~ MonthlyIncome, data = attrition) %>%
#' step_num2factor(
#' MonthlyIncome,
#' transform = binner,
#' levels = inc,
#' ordered = TRUE
#' ) %>%
#' prep()
#'
#' encoded <- bake(rec, new_data = NULL)
#'
#' table(encoded$MonthlyIncome, binner(attrition$MonthlyIncome))
#'
#' # What happens when a value is out of range?
#' ceo <- attrition %>%
#' slice(1) %>%
#' mutate(MonthlyIncome = 10^10)
#'
#' bake(rec, ceo)
step_num2factor <-
function(recipe,
...,
role = NA,
transform = function(x) x,
trained = FALSE,
levels,
ordered = FALSE,
skip = FALSE,
id = rand_id("num2factor")) {
if (!is_tune(ordered)) {
if (!is.logical(ordered) || length(ordered) != 1) {
rlang::abort("`ordered` should be a single logical variable")
}
}
if (rlang::is_missing(levels) || !is.character(levels)) {
rlang::abort("Please provide a character vector of appropriate length for `levels`.")
}
add_step(
recipe,
step_num2factor_new(
terms = enquos(...),
role = role,
transform = transform,
trained = trained,
levels = levels,
ordered = ordered,
skip = skip,
id = id
)
)
}
step_num2factor_new <-
function(terms, role, transform, trained, levels, ordered, skip, id) {
step(
subclass = "num2factor",
terms = terms,
role = role,
transform = transform,
trained = trained,
levels = levels,
ordered = ordered,
skip = skip,
id = id
)
}
get_ord_lvls_num <- function(x, foo) {
sort(unique(as.character(foo(x))))
}
#' @export
prep.step_num2factor <- function(x, training, info = NULL, ...) {
col_names <- recipes_eval_select(x$terms, training, info)
check_type(training[, col_names], types = c("double", "integer"))
res <- lapply(training[, col_names], get_ord_lvls_num, foo = x$transform)
res <- c(res, ..levels = list(x$levels))
ord <- rep(x$ordered, length(col_names))
names(ord) <- col_names
step_num2factor_new(
terms = x$terms,
role = x$role,
transform = x$transform,
trained = TRUE,
levels = res,
ordered = ord,
skip = x$skip,
id = x$id
)
}
make_factor_num <- function(x, lvl, ord, foo) {
y <- foo(x)
if (!is.integer(y)) {
y <- as.integer(y)
}
factor(lvl[y], levels = lvl, ordered = ord)
}
#' @export
bake.step_num2factor <- function(object, new_data, ...) {
col_names <- names(object$ordered)
check_new_data(col_names, object, new_data)
lvls <- object$levels[names(object$levels) == "..levels"]
object$levels <- object$levels[names(object$levels) != "..levels"]
new_data[, col_names] <-
map(new_data[, col_names],
make_factor_num,
lvl = lvls[[1]],
ord = object$ordered[1],
foo = object$transform
)
new_data
}
print.step_num2factor <-
function(x, width = max(20, options()$width - 30), ...) {
title <- "Factor variables from "
print_step(names(x$ordered), x$terms, x$trained, title, width)
invisible(x)
}
#' @rdname tidy.recipe
#' @export
tidy.step_num2factor <- function(x, ...) {
term_names <- sel2char(x$terms)
p <- length(term_names)
if (is_trained(x)) {
res <- tibble(
terms = term_names,
ordered = rep(unname(x$ordered), p)
)
} else {
res <- tibble(
terms = term_names,
ordered = rep(unname(x$ordered), p)
)
}
res$id <- x$id
res
}