tests/testthat/_snaps/misc.md

In themis: Extra Recipes Steps for Dealing with Unbalanced Data

nn_indices() errors informatively for singular mahalanobis covariance

Code
  nn_indices(data, 1, "mahalanobis")
Condition
  Error in `nn_indices()`:
  ! `distance = "mahalanobis"` requires more observations than predictors in each class.
  i 4 observations were found but 5 predictors are present.
  i Try a different `distance` metric or reduce the number of predictors.

mahalanobis errors informatively for collinear predictors (#246)

Code
  nn_indices(data, 1, "mahalanobis")
Condition
  Error in `nn_indices()`:
  ! `distance = "mahalanobis"` requires an invertible covariance matrix, but the covariance of the predictors is singular.
  i This happens when predictors are collinear or constant, or when duplicated rows leave too few distinct observations.
  i Try a different `distance` metric or remove the redundant predictors.

Code
  nn_indices(data, 1, "mahalanobis")
Condition
  Error in `nn_indices()`:
  ! `distance = "mahalanobis"` requires an invertible covariance matrix, but the covariance of the predictors is singular.
  i This happens when predictors are collinear or constant, or when duplicated rows leave too few distinct observations.
  i Try a different `distance` metric or remove the redundant predictors.

Code
  nn_indices(data, 1, "mahalanobis")
Condition
  Error in `nn_indices()`:
  ! `distance = "mahalanobis"` requires an invertible covariance matrix, but the covariance of the predictors is singular.
  i This happens when predictors are collinear or constant, or when duplicated rows leave too few distinct observations.
  i Try a different `distance` metric or remove the redundant predictors.

sqrt-embedded metrics reject non-distribution predictors

Code
  nn_indices(negative, 1, "matusita")
Condition
  Error in `nn_indices()`:
  ! `distance = "matusita"` requires non-negative predictor values.
  i Negative values were found in the columns used to compute distances.
  i Each row is treated as a probability distribution by this metric.
  i Try a different `distance` metric or rescale the predictors.

Code
  nn_indices(unnormalized, 1, "hellinger")
Condition
  Error in `nn_indices()`:
  ! `distance = "hellinger"` requires each row to sum to 1.
  i 3 rows do not sum to 1.
  i Each row is treated as a probability distribution by this metric.
  i Try `distance = "matusita"` or `distance = "squared_chord"`, which do not require this.

Code
  nn_indices(unnormalized, 1, "bhattacharyya")
Condition
  Error in `nn_indices()`:
  ! `distance = "bhattacharyya"` requires each row to sum to 1.
  i 3 rows do not sum to 1.
  i Each row is treated as a probability distribution by this metric.
  i Try `distance = "matusita"` or `distance = "squared_chord"`, which do not require this.

philentropy metrics exclude similarity measures and asymmetric ones

Code
  check_distance_arg("intersection")
Condition
  Error:
  ! `distance` must be one of "euclidean", "cosine", "mahalanobis", "manhattan", "chebyshev", "squared_chord", "matusita", "hellinger", "bhattacharyya", "canberra", "soergel", "lorentzian", "jeffreys", "topsoe", "jensen-shannon", "jensen_difference", "taneja", or "kumar-johnson", not "intersection".

Code
  check_distance_arg("kullback-leibler")
Condition
  Error:
  ! `distance` must be one of "euclidean", "cosine", "mahalanobis", "manhattan", "chebyshev", "squared_chord", "matusita", "hellinger", "bhattacharyya", "canberra", "soergel", "lorentzian", "jeffreys", "topsoe", "jensen-shannon", "jensen_difference", "taneja", or "kumar-johnson", not "kullback-leibler".

philentropy metrics that divide by values reject zeros

Code
  nn_indices(with_zero, 1, "jeffreys")
Condition
  Error in `nn_indices()`:
  ! `distance = "jeffreys"` requires strictly positive predictor values.
  i Zero or negative values were found in the columns used to compute distances.
  i This metric divides by individual values, so a zero makes the distance infinite.
  i Try `distance = "jensen-shannon"` or `distance = "canberra"`, which allow zeros.

Code
  nn_indices(with_zero, 1, "taneja")
Condition
  Error in `nn_indices()`:
  ! `distance = "taneja"` requires strictly positive predictor values.
  i Zero or negative values were found in the columns used to compute distances.
  i This metric divides by individual values, so a zero makes the distance infinite.
  i Try `distance = "jensen-shannon"` or `distance = "canberra"`, which allow zeros.

Code
  nn_indices(with_zero, 1, "kumar-johnson")
Condition
  Error in `nn_indices()`:
  ! `distance = "kumar-johnson"` requires strictly positive predictor values.
  i Zero or negative values were found in the columns used to compute distances.
  i This metric divides by individual values, so a zero makes the distance infinite.
  i Try `distance = "jensen-shannon"` or `distance = "canberra"`, which allow zeros.

philentropy metrics reject negative predictors

Code
  nn_indices(negative, 1, "canberra")
Condition
  Error in `nn_indices()`:
  ! `distance = "canberra"` requires non-negative predictor values.
  i Negative values were found in the columns used to compute distances.
  i Each row is treated as a probability distribution by this metric.
  i Try a different `distance` metric or rescale the predictors.

ratio_target() handles zero-length counts

Code
  over_target(counts, c(a = 1))
Condition
  Error:
  ! `over_ratio` names must be levels of the outcome.
  x Unknown name: "a".
  i No levels were observed in the outcome.

check_ratio() rejects malformed ratios

Code
  check_ratio(-1, arg = "over_ratio")
Condition
  Error:
  ! `over_ratio` must be a number larger than or equal to 0, not the number -1.

Code
  check_ratio(c(a = 1, b = -1), arg = "over_ratio")
Condition
  Error:
  ! `over_ratio` must be larger than or equal to 0.

Code
  check_ratio(c(1, 2), arg = "over_ratio")
Condition
  Error:
  ! `over_ratio` must be a single number or a named numeric vector.
  i Every element must be named with a level of the outcome.

Code
  check_ratio(c(a = 1, 2), arg = "over_ratio")
Condition
  Error:
  ! `over_ratio` must be a single number or a named numeric vector.
  i Every element must be named with a level of the outcome.

Code
  check_ratio(c(a = 1, a = 2), arg = "over_ratio")
Condition
  Error:
  ! `over_ratio` must have unique names, but "a" is duplicated.

Code
  check_ratio(c(a = 1, b = NA), arg = "over_ratio")
Condition
  Error:
  ! `over_ratio` must be finite, not missing or infinite.

Code
  check_ratio(c(a = "1"), arg = "over_ratio")
Condition
  Error:
  ! `over_ratio` must be a single number or a named numeric vector, not a string.

ratio_target() errors on names that are not levels

Code
  over_target(counts, c(a = 1, potato = 2))
Condition
  Error:
  ! `over_ratio` names must be levels of the outcome.
  x Unknown name: "potato".
  i Available levels: "a" and "b".

Code
  under_target(counts_zero, c(c = 1))
Condition
  Error:
  ! `under_ratio` names must be levels of the outcome.
  x Unknown name: "c".
  i Available levels: "a" and "b".

check_scalar_ratio() rejects a named vector

Code
  check_scalar_ratio(c(a = 1), arg = "over_ratio")
Condition
  Error:
  ! `over_ratio` must be a single number, not a named vector.
  i Per-class ratios are not supported here because `over_ratio` scales the size of the total generated sample.

themis documentation built on Aug. 2, 2026, 9:07 a.m.