In [1]:

    

%matplotlib inline

import matplotlib as plt
import numpy as np

from autotagger.stackoverflow.preprocess import load_pickle_sklearn_format
from sklearn import cross_validation,linear_model
from sklearn.datasets import make_multilabel_classification
from sklearn.metrics import f1_score, precision_score, recall_score
from sklearn.multiclass import OneVsRestClassifier
import pickle


    

In [2]:

    

X,Y = load_pickle_sklearn_format("1GB_100_features")


    

In [4]:

    

X_train, X_test, Y_train, Y_test = cross_validation.train_test_split(X,Y,test_size=0.80, random_state=42)
X_train.shape, X_test.shape, Y_train.shape, Y_test.shape


    

    
Out[4]:





((166811, 100), (667245, 100), (166811, 100), (667245, 100))

In [6]:

    

clf = linear_model.LinearRegression()
meta_clf = OneVsRestClassifier(clf)


    

In [7]:

    

meta_clf.fit(X_train,Y_train)


    

    
Out[7]:





OneVsRestClassifier(estimator=LinearRegression(copy_X=True, fit_intercept=True, n_jobs=1, normalize=False),
          n_jobs=1)

In [8]:

    

Y_pred = meta_clf.predict(X_test)


    

In [9]:

    

# macro average refers to the average f1_score for each label
f1_score(Y_test,Y_pred, average='macro')


    

    




/home/felipe/auto-tagger/venv3/lib/python3.4/site-packages/sklearn/metrics/classification.py:1074: UndefinedMetricWarning: F-score is ill-defined and being set to 0.0 in labels with no predicted samples.
  'precision', 'predicted', average, warn_for)






    
Out[9]:





0.018744593095539479

In [11]:

    

# if we just consider the labels that have had at least one instance predicted,
# our score goes up:

label_scores = f1_score(Y_test,Y_pred,average=None)
valid_label_indices = np.nonzero(label_scores)[0]
f1_score(Y_test,Y_pred,average='macro',labels=valid_label_indices)


    

    




/home/felipe/auto-tagger/venv3/lib/python3.4/site-packages/sklearn/metrics/classification.py:1074: UndefinedMetricWarning: F-score is ill-defined and being set to 0.0 in labels with no predicted samples.
  'precision', 'predicted', average, warn_for)






    
Out[11]:





0.23430741369424352

In [12]:

    

# micro average refers to the average f1_score for each instance
f1_score(Y_test,Y_pred,average='micro')


    

    
Out[12]:





0.13040891760528536

In [ ]: