INNER CODE UNIT · Python
_mean_match_nearest_neighbors
AnotherSamWilson/miceforest · miceforest/imputation_kernel.py:585
def _mean_match_nearest_neighbors(
mean_match_candidates: int,
bachelor_preds: DataFrame,
candidate_preds: DataFrame,
candidate_values: Series,
random_state: np.random.RandomState,
hashed_seeds: Optional[np.ndarray] = None,
) -> Series:
"""
Determines the values of candidates which will be used to impute the bachelors
"""
assert mean_match_candidates > 0, "Do not use nearest_neighbors with 0 mmc."
num_bachelors = bachelor_preds.shape[0]
# balanced_tree = False fixes a recursion issue for some reason.
# https://github.com/scipy/scipy/issues/14799
kd_tree = KDTree(candidate_preds, leafsize=16, balanced_tree=False)