knn_impute {structToolbox} | R Documentation |
Applies a k-nearest neighbour approach to impute missing values.
knn_impute( neighbours = 5, sample_max = 50, feature_max = 50, by = "features", ... )
neighbours |
The number of neighbours to use for imputation. |
sample_max |
Maximum percentage of missing values in any sample. Default = 50. |
feature_max |
Maximum percentage of missing values in any feature. Default = 50. |
by |
Impute by similar "samples" or "features". Default = "features". |
... |
additional slots and values passed to struct_class |
struct object
M = knn_impute()