In [1]:

    

from sklearn.metrics import precision_score
from sklearn.metrics import confusion_matrix


    

In [2]:

    

y_true = [0, 0, 0, 0, 0, 1, 1, 1, 1, 1]
y_pred = [0, 1, 1, 1, 1, 0, 0, 0, 1, 1]


    

In [3]:

    

print(precision_score(y_true, y_pred))


    

    




0.3333333333333333

In [4]:

    

print(precision_score(y_true, y_pred, pos_label=0))


    

    




0.25

In [5]:

    

print(precision_score(y_true, y_pred, average=None))


    

    




[0.25       0.33333333]

In [6]:

    

print(precision_score(y_true, y_pred, average='macro'))


    

    




0.29166666666666663

In [7]:

    

print(precision_score(y_true, y_pred, average='micro'))


    

    




0.3

In [8]:

    

print(confusion_matrix(y_true, y_pred))


    

    




[[1 4]
 [3 2]]

In [9]:

    

print(confusion_matrix(y_true, y_pred, labels=[1, 0]))


    

    




[[2 3]
 [4 1]]

In [10]:

    

print(precision_score(y_true, y_pred, average='weighted'))


    

    




0.29166666666666663

In [11]:

    

y_true_2 = [0, 1, 1, 1, 1]
y_pred_2 = [0, 0, 0, 0, 1]


    

In [12]:

    

print(confusion_matrix(y_true_2, y_pred_2))


    

    




[[1 0]
 [3 1]]

In [13]:

    

print(confusion_matrix(y_true_2, y_pred_2, labels=[1, 0]))


    

    




[[1 3]
 [0 1]]

In [14]:

    

print(precision_score(y_true_2, y_pred_2))


    

    




1.0

In [15]:

    

print(precision_score(y_true_2, y_pred_2, pos_label=0))


    

    




0.25

In [16]:

    

print(precision_score(y_true_2, y_pred_2, average='macro'))


    

    




0.625

In [17]:

    

print(precision_score(y_true_2, y_pred_2, average='micro'))


    

    




0.4

In [18]:

    

print(precision_score(y_true_2, y_pred_2, average='weighted'))


    

    




0.85

In [19]:

    

y_true_ab = ['A', 'A', 'A', 'A', 'A', 'B', 'B', 'B', 'B', 'B']
y_pred_ab = ['A', 'B', 'B', 'B', 'B', 'A', 'A', 'A', 'B', 'B']


    

In [20]:

    

# print(precision_score(y_true_ab, y_pred_ab))
# ValueError: pos_label=1 is not a valid label: array(['A', 'B'], dtype='<U1')


    

In [21]:

    

print(precision_score(y_true_ab, y_pred_ab, pos_label='A'))


    

    




0.25