multilabel_datagenerator.py

#!/usr/bin/env python

"""
Benchmarking test with time taken for multilabel learning using Shogun's 
Structured Output (SO) learning framework.
"""

from __future__ import print_function
from __future__ import division

import numpy as np
from sklearn.datasets import make_multilabel_classification
from sklearn.multiclass import OneVsRestClassifier
from sklearn.svm import SVC
from sklearn.metrics import jaccard_similarity_score
from sklearn.preprocessing import MultiLabelBinarizer
from time import time
from os import system


def generate_svm_light_sample(x_sample, y_sample):
    return "%s %s" % (
            ','.join(map(str, y_sample)),
            ' '.join(["%d:%f" % (index+1, val)
                for index, val in enumerate(x_sample)]))

def generate_multilabel_data(output_file, X, y):
    m, _ = X.shape
    assert(m == len(y))
    for val in range(m):
        output_file.writelines("%s\n" % generate_svm_light_sample(
            X[val], y[val]))

if __name__ == '__main__':
    print(__doc__)

    X, y = make_multilabel_classification(
            n_features=1000,
            n_samples=10000,
            n_classes=10,
            n_labels=3,
            allow_unlabeled=False)

    with open("sample_main.svm", "w+") as output_file:
        generate_multilabel_data(output_file, X, y)