In [1]:

    

from sklearn.metrics import f1_score as sklearn_f1
from autotagger.evaluation import basic


    

In [2]:

    

predicted,actual = basic.build_dummy_pred_test_pair()


    

In [3]:

    

# manually calculating f1 score and averaging the results for each instance
basic.get_samples_f1(predicted,actual)


    

    
Out[3]:





0.7000000000000001

In [4]:

    

sklearn_f1(actual,predicted,average='samples')


    

    
Out[4]:





0.70000000000000007

In [5]:

    

basic.get_micro_f1(predicted,actual)


    

    
Out[5]:





0.7142857142857143

In [6]:

    

sklearn_f1(actual,predicted,average='micro')


    

    
Out[6]:





0.7142857142857143