how to make sklearn pipeline using custom model?

Question

I want to make a sklearn pipeline using the custom Artificial Neural Network I already have. I want to make pipeline in which input goes to ANN and its output goes to the sklearn.svm.SVC model and final prediction is made. So, how can I do this using sklearn pipeline?

Answer 1

Implementing a custom transformer is simple. You have to implement the fit and transform methods like below. Since your ANN is already trained (right?) the fit method has to do nothing, just return self. And the transform method has to pass the incoming data to the ANN and return its output.

from sklearn.base import BaseEstimator, TransformerMixin

class MyANNTransformer(BaseEstimator, TransformerMixin):
    def __init__(self, ann):
        self.ann = ann

    def fit(self, X, y):
        return self

    def transform(self, X)
        return self.ann.predict(X)

Now you can include that in pipelines:

from sklearn.pipeline import make_pipeline
from sklearn.svm import SVC

pipe = make_pipeline(MyANNTransformer(ann),
                     SVC())

pipe.fit(Xtrain, ytrain)
pipe.predict(Xtest)