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
|
#' Center and scale numeric data
#'
#' `step_normalize` creates a *specification* of a recipe
#' step that will normalize numeric data to have a standard
#' deviation of one and a mean of zero.
#'
#' @inheritParams step_center
#' @param ... One or more selector functions to choose which
#' variables are affected by the step. See [selections()]
#' for more details. For the `tidy` method, these are not
#' currently used.
#' @param role Not used by this step since no new variables are
#' created.
#' @param means A named numeric vector of means. This is
#' `NULL` until computed by [prep.recipe()].
#' @param sds A named numeric vector of standard deviations This
#' is `NULL` until computed by [prep.recipe()].
#' @param na_rm A logical value indicating whether `NA`
#' values should be removed when computing the standard deviation and mean.
#' @return An updated version of `recipe` with the new step
#' added to the sequence of existing steps (if any). For the
#' `tidy` method, a tibble with columns `terms` (the
#' selectors or variables selected), `value` (the
#' standard deviations and means), and `statistic` for the type of value.
#' @keywords datagen
#' @concept preprocessing
#' @concept normalization_methods
#' @export
#' @details Centering data means that the average of a variable is subtracted
#' from the data. Scaling data means that the standard deviation of a variable
#' is divided out of the data. `step_normalize` estimates the variable standard
#' deviations and means from the data used in the `training` argument of
#' `prep.recipe`. `bake.recipe` then applies the scaling to new data sets using
#' these estimates.
#' @examples
#' library(modeldata)
#' data(biomass)
#'
#' biomass_tr <- biomass[biomass$dataset == "Training",]
#' biomass_te <- biomass[biomass$dataset == "Testing",]
#'
#' rec <- recipe(HHV ~ carbon + hydrogen + oxygen + nitrogen + sulfur,
#' data = biomass_tr)
#'
#' norm_trans <- rec %>%
#' step_normalize(carbon, hydrogen)
#'
#' norm_obj <- prep(norm_trans, training = biomass_tr)
#'
#' transformed_te <- bake(norm_obj, biomass_te)
#'
#' biomass_te[1:10, names(transformed_te)]
#' transformed_te
#' tidy(norm_trans, number = 1)
#' tidy(norm_obj, number = 1)
#'
step_normalize <-
function(recipe,
...,
role = NA,
trained = FALSE,
means = NULL,
sds = NULL,
na_rm = TRUE,
skip = FALSE,
id = rand_id("normalize")) {
add_step(
recipe,
step_normalize_new(
terms = ellipse_check(...),
role = role,
trained = trained,
means = means,
sds = sds,
na_rm = na_rm,
skip = skip,
id = id
)
)
}
step_normalize_new <-
function(terms, role, trained, means, sds, na_rm, skip, id) {
step(
subclass = "normalize",
terms = terms,
role = role,
trained = trained,
means = means,
sds = sds,
na_rm = na_rm,
skip = skip,
id = id
)
}
#' @export
prep.step_normalize <- function(x, training, info = NULL, ...) {
col_names <- eval_select_recipes(x$terms, training, info)
check_type(training[, col_names])
means <- vapply(training[, col_names], mean, c(mean = 0), na.rm = x$na_rm)
sds <- vapply(training[, col_names], sd, c(sd = 0), na.rm = x$na_rm)
step_normalize_new(
terms = x$terms,
role = x$role,
trained = TRUE,
means = means,
sds = sds,
na_rm = x$na_rm,
skip = x$skip,
id = x$id
)
}
#' @export
bake.step_normalize <- function(object, new_data, ...) {
res <- sweep(as.matrix(new_data[, names(object$means)]), 2, object$means, "-")
res <- sweep(res, 2, object$sds, "/")
res <- tibble::as_tibble(res)
new_data[, names(object$sds)] <- res
as_tibble(new_data)
}
print.step_normalize <-
function(x, width = max(20, options()$width - 30), ...) {
cat("Centering and scaling for ", sep = "")
printer(names(x$sds), x$terms, x$trained, width = width)
invisible(x)
}
#' @rdname step_normalize
#' @param x A `step_normalize` object.
#' @export
tidy.step_normalize <- function(x, ...) {
if (is_trained(x)) {
res <- tibble(terms = c(names(x$means), names(x$sds)),
statistic = rep(c("mean", "sd"), each = length(x$sds)),
value = c(x$means, x$sds))
} else {
term_names <- sel2char(x$terms)
res <- tibble(terms = term_names,
statistic = na_chr,
value = na_dbl)
}
res$id <- x$id
res
}
|
