Performs a normalization of data, i.e., it scales variables in the range
0 - 1. This is a special case of data_rescale()
. unnormalize()
is the
counterpart, but only works for variables that have been normalized with
normalize()
.
Usage
normalize(x, ...)
# S3 method for numeric
normalize(x, include_bounds = TRUE, verbose = TRUE, ...)
# S3 method for data.frame
normalize(
x,
select = NULL,
exclude = NULL,
include_bounds = TRUE,
ignore_case = FALSE,
verbose = TRUE,
...
)
unnormalize(x, ...)
# S3 method for numeric
unnormalize(x, verbose = TRUE, ...)
# S3 method for data.frame
unnormalize(
x,
select = NULL,
exclude = NULL,
ignore_case = FALSE,
verbose = TRUE,
...
)
Arguments
- x
A numeric vector, (grouped) data frame, or matrix. See 'Details'.
- ...
Arguments passed to or from other methods.
- include_bounds
Logical, if
TRUE
, return value may include 0 and 1. IfFALSE
, the return value is compressed, using Smithson and Verkuilen's (2006) formula(x * (n - 1) + 0.5) / n
, to avoid zeros and ones in the normalized variables. This can be useful in case of beta-regression, where the response variable is not allowed to include zeros and ones.- verbose
Toggle warnings and messages on or off.
- select
Variables that will be included when performing the required tasks. Can be either
a variable specified as a literal variable name (e.g.,
column_name
),a string with the variable name (e.g.,
"column_name"
), or a character vector of variable names (e.g.,c("col1", "col2", "col3")
),a formula with variable names (e.g.,
~column_1 + column_2
),a vector of positive integers, giving the positions counting from the left (e.g.
1
orc(1, 3, 5)
),a vector of negative integers, giving the positions counting from the right (e.g.,
-1
or-1:-3
),one of the following select-helpers:
starts_with("")
,ends_with("")
,contains("")
, a range using:
orregex("")
,or a function testing for logical conditions, e.g.
is.numeric()
(oris.numeric
), or any user-defined function that selects the variables for which the function returnsTRUE
(like:foo <- function(x) mean(x) > 3
),ranges specified via literal variable names, select-helpers (except
regex()
) and (user-defined) functions can be negated, i.e. return non-matching elements, when prefixed with a-
, e.g.-ends_with("")
,-is.numeric
or-Sepal.Width:Petal.Length
. Note: Negation means that matches are excluded, and thus, theexclude
argument can be used alternatively. For instance,select=-ends_with("Length")
(with-
) is equivalent toexclude=ends_with("Length")
(no-
). In case negation should not work as expected, use theexclude
argument instead.
If
NULL
, selects all columns. Patterns that found no matches are silently ignored, e.g.find_columns(iris, select = c("Species", "Test"))
will just return"Species"
.- exclude
See
select
, however, column names matched by the pattern fromexclude
will be excluded instead of selected. IfNULL
(the default), excludes no columns.- ignore_case
Logical, if
TRUE
and when one of the select-helpers or a regular expression is used inselect
, ignores lower/upper case in the search pattern when matching against variable names.
Details
If
x
is a matrix, normalization is performed across all values (not column- or row-wise). For column-wise normalization, convert the matrix to a data.frame.If
x
is a grouped data frame (grouped_df
), normalization is performed separately for each group.
Selection of variables - the select
argument
For most functions that have a select
argument (including this function),
the complete input data frame is returned, even when select
only selects
a range of variables. That is, the function is only applied to those variables
that have a match in select
, while all other variables remain unchanged.
In other words: for this function, select
will not omit any non-included
variables, so that the returned data frame will include all variables
from the input data frame.
References
Smithson M, Verkuilen J (2006). A Better Lemon Squeezer? Maximum-Likelihood Regression with Beta-Distributed Dependent Variables. Psychological Methods, 11(1), 54–71.
See also
Other transform utilities:
data_rescale()
,
data_reverse()
,
ranktransform()
,
standardize()
Examples
normalize(c(0, 1, 5, -5, -2))
#> [1] 0.5 0.6 1.0 0.0 0.3
#> attr(,"include_bounds")
#> [1] TRUE
#> attr(,"min_value")
#> [1] -5
#> attr(,"range_difference")
#> [1] 10
normalize(c(0, 1, 5, -5, -2), include_bounds = FALSE)
#> [1] 0.50 0.58 0.90 0.10 0.34
#> attr(,"include_bounds")
#> [1] FALSE
#> attr(,"min_value")
#> [1] -5
#> attr(,"range_difference")
#> [1] 10
head(normalize(trees))
#> Girth Height Volume
#> 1 0.00000000 0.29166667 0.001497006
#> 2 0.02439024 0.08333333 0.001497006
#> 3 0.04065041 0.00000000 0.000000000
#> 4 0.17886179 0.37500000 0.092814371
#> 5 0.19512195 0.75000000 0.128742515
#> 6 0.20325203 0.83333333 0.142215569