how to change estimator in voting classifier?

Question

I need to do voting classifier to my models. Apparently, i cannot do it since my models are not part of scikit learn so i cannot do voting classifier. So i decide that i have to fit my models first then do voting classifier but unfortunately voting classifier doesn't support prefit models. How to change estimator so that voting classifier works on prefit model? I work on python 2.7

This is my code:

import numpy as np
import os
from decimal import *
import sys
sys.path.append('/home/nida/pylibol/python')
import classifiers
from classifiers_new2 import *
from sklearn.datasets import load_svmlight_file

filedir= '/home/nida/2018/2018200.libsvm'
X, y = load_svmlight_file(filedir)

print 'X shape:\n',X.shape
print 'y shape:\n',y.shape

X_train, X_test, y_train, y_test = train_test_split(X, y, train_size=0.8, random_state=0)

print 'X_test shape:\n',X_test.shape
print 'y_test shape:\n',y_test.shape
print 'X train:\n',X_train.shape
print 'y train:\n',y_train.shape

model1 = Pa1()
model2 = Ogd()
model3= Arow()

model1.fit(X_train,y_train)
model2.fit(X_train,y_train)
model3.fit(X_train,y_train)

pred1=model1.predict(X_test)
pred2=model2.predict(X_test)
pred3=model3.predict(X_test)

modelsNames = [('Pa1',model1), ('OGD',model2),('Arow',model3)]

votingClassifier = VotingClassifier(voting = 'hard',estimators= modelsNames)
votingClassifier.fit(X_train,y_train)

This is error:

TypeError                                 Traceback (most recent call last)
<ipython-input-22-6f0e98ffe925> in <module>()
     36 
     37 votingClassifier = VotingClassifier(voting = 'hard',estimators= modelsNames)
---> 38 votingClassifier.fit(X_train,y_train)

/home/nida/anaconda3/envs/py2/lib/python2.7/site-packages/sklearn/ensemble/voting_classifier.pyc in fit(self, X, y, sample_weight)
    198             delayed(_parallel_fit_estimator)(clone(clf), X, transformed_y,
    199                                              sample_weight=sample_weight)
--> 200             for clf in clfs if clf is not None)
    201 
    202         self.named_estimators_ = Bunch(**dict())

/home/nida/anaconda3/envs/py2/lib/python2.7/site-packages/sklearn/externals/joblib/parallel.pyc in __call__(self, iterable)
    919             # remaining jobs.
    920             self._iterating = False
--> 921             if self.dispatch_one_batch(iterator):
    922                 self._iterating = self._original_iterator is not None
    923 

/home/nida/anaconda3/envs/py2/lib/python2.7/site-packages/sklearn/externals/joblib/parallel.pyc in dispatch_one_batch(self, iterator)
    752             tasks = BatchedCalls(itertools.islice(iterator, batch_size),
    753                                  self._backend.get_nested_backend(),
--> 754                                  self._pickle_cache)
    755             if len(tasks) == 0:
    756                 # No more tasks available in the iterator: tell caller to stop.

/home/nida/anaconda3/envs/py2/lib/python2.7/site-packages/sklearn/externals/joblib/parallel.pyc in __init__(self, iterator_slice, backend_and_jobs, pickle_cache)
    208 
    209     def __init__(self, iterator_slice, backend_and_jobs, pickle_cache=None):
--> 210         self.items = list(iterator_slice)
    211         self._size = len(self.items)
    212         if isinstance(backend_and_jobs, tuple):

/home/nida/anaconda3/envs/py2/lib/python2.7/site-packages/sklearn/ensemble/voting_classifier.pyc in <genexpr>((clf,))
    198             delayed(_parallel_fit_estimator)(clone(clf), X, transformed_y,
    199                                              sample_weight=sample_weight)
--> 200             for clf in clfs if clf is not None)
    201 
    202         self.named_estimators_ = Bunch(**dict())

/home/nida/anaconda3/envs/py2/lib/python2.7/site-packages/sklearn/base.pyc in clone(estimator, safe)
     58                             "it does not seem to be a scikit-learn estimator "
     59                             "as it does not implement a 'get_params' methods."
---> 60                             % (repr(estimator), type(estimator)))
     61     klass = estimator.__class__
     62     new_object_params = estimator.get_params(deep=False)

TypeError: Cannot clone object '<classifiers_new2.Pa1 instance at 0x7f1c9c0e41e0>' (type <type 'instance'>): it does not seem to be a scikit-learn estimator as it does not implement a 'get_params' methods.

Answer 1

You should implement you estimator by inherit from BaseEstimator and ClassifierMixin classes. Sklearn provides great documentation how to develop custom estimator.

You estimator should looks like:

from sklearn.base import BaseEstimator, ClassifierMixin


class MyClassifier(BaseEstimator, ClassifierMixin):
    def fit(self, X, y):
        pass

    def predict(self, X):
        pass