Copy CROUST and/or iCROUST files to you directory.

How to import:
  from CROUST import croust
  from iCROUST import icroust

How to use:
  CROUST:
    function = croust(X, Y, points_to_remove, num_clusters=10000, maj_class=0, min_class=1)
  Usage:
    data, target_variable = croust(data, target_variable, points_to_remove)
    points_to_remove are the number of samples you want to remove of the majority class
    num_clusters is the number of clusters you want to define for the data. Larger samples will usually require larger clusters. Experiment with this variable to get best results
    maj_class is the class value for the majority class
    min_class is the class value for the minority class
    
  iCROUST:
    function = icroust(X, Y, points_to_remove, n_neighbours, points_to_remove_at_a_time=4, num_clusters=10000, maj_class=0, min_class=1):
  Usage:
    data, target_variable = icroust(data, target_variable, points_to_remove, n_neighbours)
    points_to_remove are the number of samples you want to remove of the majority class
    n_neighbours are the number of neighbouring samples you want to consider
    points_to_remove_at_a_time are the number of samples you want to delete in a single iCROUST iteration. Works like a batch_size variable. Reduce this value for better accuracy at the cost of speed
    num_clusters is the number of clusters you want to define for the data. Larger samples will usually require larger clusters. Experiment with this variable to get best results
    maj_class is the class value for the majority class
    min_class is the class value for the minority class
    
The paper explaining the algorithm will be uploaded once it is published at ICDM (Industrial Conference on Data Mining)