kmeans_smote_impl.md
In themis: Extra Recipes Steps for Dealing with Unbalanced Data

errors when no cluster is suitable

Code
  kmeans_smote(df, "class", num_clusters = 3)
Condition
  Error in `kmeans_smote()`:
  ! No cluster is suitable for over-sampling the minority class "a".
  i No cluster both reached `cluster_balance_threshold = 1` and contained more than 2 observations of that class.
  i Try a smaller `cluster_balance_threshold`, a smaller `num_clusters`, or fewer `neighbors`.

errors when there are too few distinct observations

Code
  kmeans_smote(df, "class", num_clusters = 10)
Condition
  Error in `kmeans_smote()`:
  ! Not enough distinct observations to compute 10 clusters.
  i 3 distinct observations were found.
  i Try a smaller `num_clusters` or remove duplicated rows.

bad args

Code
  kmeans_smote(circle_numeric, "x")
Condition
  Error in `kmeans_smote()`:
  ! `x` should refer to a factor or character column, not a double vector.

Code
  kmeans_smote(circle_example, "class")
Condition
  Error in `kmeans_smote()`:
  ! All columns for this function should be numeric. Non-numeric column found: `id`.

Code
  kmeans_smote(circle_numeric, "class", k = "yes")
Condition
  Error in `kmeans_smote()`:
  ! `k` must be a whole number, not the string "yes".

Code
  kmeans_smote(circle_numeric, "class", over_ratio = "yes")
Condition
  Error in `kmeans_smote()`:
  ! `over_ratio` must be a number, not the string "yes".

Code
  kmeans_smote(circle_numeric, "class", num_clusters = 1)
Condition
  Error in `kmeans_smote()`:
  ! `num_clusters` must be a whole number larger than or equal to 2 or `NULL`, not the number 1.

Code
  kmeans_smote(circle_numeric, "class", cluster_balance_threshold = Inf)
Condition
  Error in `kmeans_smote()`:
  ! `cluster_balance_threshold` must be a number, not `Inf`.

Code
  kmeans_smote(circle_numeric, "class", density_exponent = -1)
Condition
  Error in `kmeans_smote()`:
  ! `density_exponent` must be a number larger than or equal to 0 or `NULL`, not the number -1.

Code
  kmeans_smote(circle_numeric, "class", distance = "L2")
Condition
  Error in `kmeans_smote()`:
  ! `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 "L2".