Source code for skmultilearn.ensemble.fixed

from .partition import LabelSpacePartitioningClassifier
import copy
import random
import numpy as np
from scipy import sparse


[docs]class FixedLabelPartitionClassifier(LabelSpacePartitioningClassifier): """Classify given a fixed Label Space partition""" def __init__(self, classifier=None, require_dense=None, partition=None): super(FixedLabelPartitionClassifier, self).__init__( classifier=classifier, require_dense=require_dense) self.partition = partition self.copyable_attrs = ['partition', 'classifier', 'require_dense']
[docs] def generate_partition(self, X, y): """Assign fixed partition of the label space Mock function, the partition is assigned in the constructor. :param X: not used, maintained for api compatibility :param y: binary indicator matrix with label assignments :type y: dense or sparse matrix of {0, 1} (n_samples, n_labels) Sets `self.model_count` and `self.label_count`. """ self.label_count = y.shape[1] self.model_count = len(self.partition)