dialnd/imbalanced-algorithms

Python-based implementations of algorithms for learning on imbalanced data.

imbalanced-algorithms's Issues

In gan.py pre-train = True leads to error on Python 2.7 and 3.6

The error message is:
AttributeError: 'mtrand.RandomState' object has no attribute 'random'

The fix is:
Change mtrand to np.random.rand.

is:issue is:open Multi-class SMOTEBoost

Just curious to ask, does this package supports a multi-class problem with a multi-minority scenario? Not sure if I read this correctly, but it seems to me the support is only for the binary class problem:

if minority_target is None:
    # Determine the minority class label.
    stats_c_ = Counter(y)
    maj_c_ = max(stats_c_, key=stats_c_.get)
    min_c_ = min(stats_c_, key=stats_c_.get)
    self.minority_target = min_c_
else:
    self.minority_target = minority_target

In my current task, I have a multi-majority multi-minority scenario:

Class = 0,	Count = 18749,	Percentage = 22.01
Class = 1,	Count = 3482,	Percentage = 4.09
Class = 2,	Count = 9566,	Percentage = 11.23
Class = 3,	Count = 49741,	Percentage = 58.4
Class = 4,	Count = 3634,	Percentage = 4.27

Toy example

Thank you for sharing the code. Could you also share a toy example which make use of these algorithms?