sjmisc (version 1.1)

set_na: Set NA for specific variable values

Description

This function sets specific values of a variable, data frame or list of variables as missings (NA).

Usage

set_na(x, values, as.attr = FALSE)

Arguments

x

Variable (vector), data.frame or list of variables where new missing values should be defined. If x is a data.frame, each column is assumed to be a new variable, where missings should be defined.

values

Numeric vector with values that should be replaced with NA's. Thus, for each variable in x, values are replaced by NA's. Or: a logical vector describing which values should

as.attr

Logical, if TRUE, values of x will not be converted to NA. Rather, the missing code values of values will be added as missing-attribute is_na

Value

• x, where each value of values is replaced by an NA.

Details

set_na converts those values to NA that are specified in the function's values argument; hence, by default, set_na ignores any missing code attributes like is_na. to_na, by contrast, converts values to NA, which are defined as missing through the is_na-attribute of a vector (see labelled). If as.attr = TRUE, values in x will not be converted to NA. Instead, the attribute is_na will be added to x, indicating which values should be coded as missing. values may either be numeric, with each number indicating a value that should be defined as missing; or a vector of logicals, describing which values should be translated to missing values (see 'Examples'). Furthermore, see 'Details' in get_na.

See Also

replace_na to replace NA's with specific values, rec for general recoding of variables and recode_to for re-shifting value ranges. See get_na to get values of missing values in labelled vectors and to_na to convert missing value codes into NA.

Examples

# create random variable
dummy <- sample(1:8, 100, replace = TRUE)
# show value distribution
table(dummy)
# set value 1 and 8 as missings
dummy <- set_na(dummy, c(1, 8))
# show value distribution, including missings
table(dummy, exclude = NULL)

# create sample data frame
dummy <- data.frame(var1 = sample(1:8, 100, replace = TRUE),
                    var2 = sample(1:10, 100, replace = TRUE),
                    var3 = sample(1:6, 100, replace = TRUE))
# show head of data frame
head(dummy)
# set value 2 and 4 as missings
dummy <- set_na(dummy, c(2, 4))
# show head of new data frame
head(dummy)

# create list of variables
data(efc)
dummy <- list(efc$c82cop1, efc$c83cop2, efc$c84cop3)
# check original distribution of categories
lapply(dummy, table, exclude = NULL)
# set 3 to NA
lapply(set_na(dummy, 3), table, exclude = NULL)

# create random variable
dummy <- sample(1:5, 100, replace = TRUE)
# declare missing values, but only as attribute
dummy <- set_na(dummy, c(3, 5), as.attr = TRUE)

str(dummy)
table(dummy)
get_na(dummy)

# create random variable
dummy <- sample(1:5, 100, replace = TRUE)
# declare missing values, but only as attribute
# missing code definition may be logical indices
dummy <- set_na(dummy,
                c(FALSE, FALSE, FALSE, TRUE, TRUE),
                as.attr = TRUE)

str(dummy)
table(dummy)
get_na(dummy)