Additional Assertions for checkmate

Additional assertion functions which can be used together with the checkmate package.

Usage

assert_list_of_variables(x, .var.name = checkmate::vname(x), add = NULL)

assert_df_with_variables(
  df,
  variables,
  na_level = NULL,
  .var.name = checkmate::vname(df),
  add = NULL
)

assert_valid_factor(
  x,
  min.levels = 1,
  max.levels = NULL,
  null.ok = TRUE,
  any.missing = TRUE,
  n.levels = NULL,
  len = NULL,
  .var.name = checkmate::vname(x),
  add = NULL
)

assert_df_with_factors(
  df,
  variables,
  min.levels = 1,
  max.levels = NULL,
  any.missing = TRUE,
  na_level = NULL,
  .var.name = checkmate::vname(df),
  add = NULL
)

assert_proportion_value(x, include_boundaries = FALSE)

Arguments

x: (any)
object to test.
.var.name: [character(1)]
Name of the checked object to print in assertions. Defaults to the heuristic implemented in vname.
add: [AssertCollection]
Collection to store assertion messages. See AssertCollection.
df: (data frame)
data set to test.
variables: (named list of character)
list of variables to test.
na_level: (character)
the string you have been using to represent NA or missing data. For NA values please consider using directly base::is.na or similar approaches.
min.levels: [integer(1)]
Minimum number of factor levels. Default is NULL (no check).
max.levels: [integer(1)]
Maximum number of factor levels. Default is NULL (no check).
null.ok: [logical(1)]
If set to TRUE, x may also be NULL. In this case only a type check of x is performed, all additional checks are disabled.
any.missing: [logical(1)]
Are vectors with missing values allowed? Default is TRUE.
n.levels: [integer(1)]
Exact number of factor levels. Default is NULL (no check).
len: [integer(1)]
Exact expected length of x.
include_boundaries: (logical)
whether to include boundaries when testing for proportions.
(`integer`): minimum number of factor levels. Default is 1.
...: a collection of objects to test.

Functions

assert_list_of_variables(): Check whether x is a valid list of variable names. NULL elements of the list x are dropped out with Filter(Negate(is.null), x).
assert_df_with_variables(): Check whether df is a data frame with the analysis variables. Please notice how this produces an error when not all variables are present in the data.frame while the opposite is not required.
assert_valid_factor(): Check whether x is a valid factor (has levels and no empty string levels). Note that NULL and NA elements are allowed.
assert_df_with_factors(): Check whether df is a data frame where the analysis variables are all factors. Note that the creation of NA by direct call of factor() will trim NA levels out of the vector list itself.
assert_proportion_value(): Check whether x is a proportion: number between 0 and 1.

Examples

# Check whether `x` is a valid list of variable names.

# Internal function - assert_list_of_variables
if (FALSE) {
assert_list_of_variables(list(val = "a"))
assert_list_of_variables(list(val = c("a", "b")))
assert_list_of_variables(list(val = c("a", "b"), val2 = NULL))

# The following calls fail
assert_list_of_variables(list(1, 2))
assert_list_of_variables(list("bla" = 2))
}

# Check whether `df` contains the analysis `variables`.

# Internal function - assert_df_with_variables
if (FALSE) {
assert_df_with_variables(
  df = data.frame(a = 5, b = 3),
  variables = list(val = "a")
)
assert_df_with_variables(
  df = data.frame(a = 5, b = 3),
  variables = list(val = c("a", "b"))
)
assert_df_with_variables(
  df = data.frame(a = 5, b = 3),
  variables = list(val = c("a", "b"))
)
assert_df_with_variables(
  df = data.frame(a = 5, b = 3, e = "<Missing>"),
  variables = list(val = c("a", "b")), na_level = "<Missing>"
)

# The following calls fail
assert_df_with_variables(
  df = matrix(1:5, ncol = 2, nrow = 3),
  variables = list(val = "a")
)
assert_df_with_variables(
  df = data.frame(a = 5, b = 3),
  variables = list(val = c("a", "b", "c"))
)
assert_df_with_variables(
  df = data.frame(a = 5, b = 3, e = "<Missing>"),
  variables = list(val = c("a", "b", "e")), na_level = "<Missing>"
)
}

# Check whether `x` is a valid factor.

# Internal function - assert_valid_factor
if (FALSE) {
assert_valid_factor(factor(c("a", NULL)))
assert_valid_factor(factor(c("a", "b")))
assert_valid_factor(factor(c("a", "b")), len = 2)
assert_valid_factor(factor(c("a", NA)), any.missing = TRUE)
assert_valid_factor(factor("A", levels = c("A", "B")))

# The following calls fail
assert_valid_factor(-1)
assert_valid_factor(factor(c("a", "")))
assert_valid_factor(factor(c("a", NA)), any.missing = FALSE)
assert_valid_factor(factor(NULL))
assert_valid_factor(factor(c(NULL, "")))
assert_valid_factor(factor())
}

# Check whether `df` contains all factor analysis `variables`.
adf <- data.frame(a = factor(c("A", "B")), b = 3)
bdf <- data.frame(a = factor(letters[1:3]), b = factor(c(1, 2, 3)), d = 3)

# Internal function - assert_df_with_factors
if (FALSE) {
assert_df_with_factors(df = adf, variables = list(val = "a"))
assert_df_with_factors(df = adf, variables = list(val = "a"), min.levels = 1)
assert_df_with_factors(df = adf, variables = list(val = "a"), min.levels = 2, max.levels = 2)
assert_df_with_factors(
  df = data.frame(a = factor(c("A", NA, "B")), b = 3),
  variable = list(val = "a"),
  min.levels = 2,
  max.levels = 2
)

# The following calls fail
assert_df_with_factors(df = adf, variables = list(val = "a"), min.levels = 1, max.levels = 1)
assert_df_with_factors(df = adf, variables = list(val = "a"), min.levels = 1, max.levels = 1)
assert_df_with_factors(df = adf, variables = list(val = "a", val = "b", val = ""))
assert_df_with_factors(df = adf, variables = list(val = "a", val = "b", val = "d"))
assert_df_with_factors(
  df = bdf,
  variables = list(val = "a", val = "b"),
  min.levels = 1,
  max.levels = 1
)
}

# Check whether `x` is between 0 and 1.
# Internal function - assert_proportion_value
if (FALSE) {
assert_proportion_value(x = 0, include_boundaries = TRUE)
assert_proportion_value(x = 0.3)

# These fail
assert_proportion_value(x = 1.3)
assert_proportion_value(x = 1)
}