Decision_function expects one argument

[samuelazran]

First of all, this is a super cool and interesting algorithm! thank you for sharing it :)

The code fails on

def decision_function(self, X): 
return self.classifier.decision_function(self, X)

with an error stating that decision_function expects only one argument.

Shouldn't it be self.classifier.decision_function(X) without the self?

Cheers,

[xuanli19]

I think so , the "self" should be remove , than the code run successful
My code

[xuanli19]

I found there is a serious mistake ...
in the code of AugBoostClassifier predict funcition :
def predict(self, X): return self.classifier.predict(X)
only use the GBDT with origin feature to predict , instead of using origin feature with new feature ..