code cleanup

multi_imbalance/resampling/ccr.py

@@ -1,5 +1,5 @@
 from collections import Counter
-from typing import Tuple, List
+from typing import Tuple, List, Callable
 
 import numpy as np
 from imblearn.base import BaseSampler
@@ -21,14 +21,18 @@ class CCR(BaseSampler):
  International Journal of Applied Mathematics and Computer Science 2017
  """
 
- def __init__(self, energy: float):
+ def __init__(self, energy: float, distance_function: Callable[[np.ndarray, np.ndarray], np.ndarray] =
+ lambda x, y: np.linalg.norm(x - y, ord=1, axis=1)) -> None:
  """
  :param energy:
  initial energy budget for each minority example to use for sphere expansion
+ :param distance_function:
+ function to calculate distance between minority example and array of majority examples, defaults to L1 norm
  """
  super().__init__()
  self.energy = energy
  self._sampling_type = "over-sampling"
+ self.distance_function = distance_function
 
  def _fit_resample(self, X: np.ndarray, y: np.ndarray) -> Tuple[np.ndarray, np.ndarray]:
  """
@@ -64,13 +68,21 @@ def _clean_and_generate(self, minority_examples: np.ndarray, majority_examples:
  :return:
  clean majority X, synthetic minority X
  """
- majority_count = len(majority_examples)
+ r, t = self._calculate_radius_and_translations(minority_examples, majority_examples)
+ translated_majority_examples = majority_examples + t
+ synthetic_examples = self._generate_synthetic_examples(minority_examples, majority_examples, r,
+ synthetic_examples_total)
+
+ return translated_majority_examples, synthetic_examples
 
+ def _calculate_radius_and_translations(self, minority_examples, majority_examples):
  r = np.zeros(minority_examples.shape[0])
  t = np.zeros(majority_examples.shape)
 
- for i, x in enumerate(minority_examples):
- distances = self.distances(x, majority_examples)
+ majority_count = len(majority_examples)
+ for i, minority_example in enumerate(minority_examples):
+ distances = self.distance_function(minority_example, majority_examples)
+
  sorted_distances_index = np.argsort(distances)
  energy = self.energy
  current_example = 0
@@ -100,18 +112,18 @@ def _clean_and_generate(self, minority_examples: np.ndarray, majority_examples:
  d = distances[j]
  if d == 0:
  continue
- translation = majority_examples[j] - x
+ translation = majority_examples[j] - minority_example
  t[j] += (r[i] - d) / d * translation
+ return r, t
 
- majority_examples += t
-
+ def _generate_synthetic_examples(self, majority_examples, minority_examples, r, synthetic_examples_total):
  generation_order = r.argsort()
  if synthetic_examples_total is None:
  synthetic_examples_total = majority_examples.shape[0] - minority_examples.shape[0]
+
  synthetic_examples_counts = (r ** -1 / (r ** -1).sum()) * synthetic_examples_total
- synthetic_leftovers = int((synthetic_examples_counts - synthetic_examples_counts.astype(int)).sum())
  synthetic_examples_counts = np.floor(synthetic_examples_counts).astype(int)
-
+ synthetic_leftovers = int((synthetic_examples_counts - synthetic_examples_counts.astype(int)).sum())
  for i in range(synthetic_leftovers):
  synthetic_examples_counts[generation_order[i % len(generation_order)]] += 1
 
@@ -134,10 +146,7 @@ def _clean_and_generate(self, minority_examples: np.ndarray, majority_examples:
  else:
  generated = np.empty((0, minority_examples.shape[1]))
 
- return majority_examples, generated
-
- def distances(self, minority_example: np.ndarray, majority_examples: np.ndarray) -> np.ndarray:
- return (abs(minority_example - majority_examples)).sum(1)
+ return generated
 
 
 class MultiClassCCR(BaseSampler):