

In [1]:

    
import sys
sys.path.append('../src/mane/prototype/')
import numpy as np
import graph as g
import pickle as p

from sklearn.preprocessing import normalize, scale, MultiLabelBinarizer
from sklearn.metrics import f1_score
from sklearn.multiclass import OneVsRestClassifier
from sklearn.svm import LinearSVC
from sklearn.linear_model import LogisticRegression
from sklearn.linear_model import LogisticRegressionCV

Experiment with Random-walk result



In [2]:

    
# Load weight
with open('../src/mane/prototype/embeddings/BC3047.weights', 'rb') as f:
    w = p.load(f)
# Load graph
bc = g.graph_from_pickle('../src/mane/data/blogcatalog3.graph', 
                         '../src/mane/data/blogcatalog3.community')



In [3]:

    
emb = (w[0] + w[1]) / 2
emb = normalize(emb)



In [4]:

    
x_train, yl_train, x_test, yl_test = bc.get_ids_labels(0.5)



In [5]:

    
lg = OneVsRestClassifier(LogisticRegression())



In [6]:

    
X_train = [emb[i] for i in x_train]
Y_train = MultiLabelBinarizer(classes=range(1,40)).fit_transform(yl_train)



In [7]:

    
lg.fit(X_train, Y_train)









    Out[7]:





OneVsRestClassifier(estimator=LogisticRegression(C=1.0, class_weight=None, dual=False, fit_intercept=True,
          intercept_scaling=1, max_iter=100, multi_class='ovr', n_jobs=1,
          penalty='l2', random_state=None, solver='liblinear', tol=0.0001,
          verbose=0, warm_start=False),
          n_jobs=1)



In [8]:

    
lg.predict_proba(emb[9566].reshape(1,-1)).shape









    Out[8]:





(1, 39)



In [9]:

    
X_test = [emb[i] for i in x_test]
Y_test = MultiLabelBinarizer(classes=range(1,40)).fit_transform(yl_test)



In [10]:

    
pred = lg.predict_proba([i for i in X_test])



In [11]:

    
Y_test









    Out[11]:





array([[0, 0, 0, ..., 0, 0, 0],
       [0, 0, 0, ..., 0, 0, 0],
       [0, 0, 0, ..., 0, 0, 0],
       ..., 
       [0, 0, 0, ..., 0, 0, 0],
       [0, 0, 0, ..., 0, 0, 0],
       [0, 1, 0, ..., 0, 0, 0]])

Get top k prediction



In [35]:

    
pred[0].argsort()[-3:]









    Out[35]:





array([5, 6, 4])



In [12]:

    
len(pred)









    Out[12]:





5156



In [13]:

    
len(yl_test)









    Out[13]:





5156



In [14]:

    
pred[0]









    Out[14]:





array([ 0.00398161,  0.0357896 ,  0.08584289,  0.02449758,  0.15023545,
        0.48816811,  0.02156142,  0.05960207,  0.06594667,  0.00866716,
        0.05512942,  0.0033318 ,  0.04651478,  0.03337085,  0.01032531,
        0.02674524,  0.04656247,  0.02798097,  0.08513218,  0.00963393,
        0.0087078 ,  0.00683082,  0.0105208 ,  0.25379662,  0.01822022,
        0.01270774,  0.00683734,  0.02172979,  0.01812818,  0.02239905,
        0.01263985,  0.02685956,  0.00782481,  0.00602085,  0.00532535,
        0.01669904,  0.00731791,  0.00314692,  0.0011479 ])



In [15]:

    
yl_test[0]









    Out[15]:





[6]



In [16]:

    
pred[0].argsort()[-1:]









    Out[16]:





array([5])



In [17]:

    
num_pred = []



In [18]:

    
for i, j in enumerate(pred):
    k = len(yl_test[i])
    num_pred.append(j.argsort()[-k:])



In [19]:

    
len(num_pred)









    Out[19]:





5156



In [20]:

    
num_pred[0]









    Out[20]:





array([5])



In [21]:

    
num_pred[1]









    Out[21]:





array([19,  6])



In [22]:

    
Y_pred = MultiLabelBinarizer(classes=range(1,40)).fit_transform(num_pred)



In [23]:

    
Y_pred[0]









    Out[23]:





array([0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
       0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0])



In [24]:

    
Y_test[0]









    Out[24]:





array([0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
       0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0])



In [25]:

    
for i, j in enumerate(num_pred):
    num_pred[i] = [k+1 for k in j]



In [26]:

    
num_pred[0]









    Out[26]:





[6]



In [27]:

    
yl_test[0]









    Out[27]:





[6]



In [28]:

    
Y_pred = MultiLabelBinarizer(classes=range(1,40)).fit_transform(num_pred)



In [29]:

    
Y_pred[0]









    Out[29]:





array([0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
       0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0])



In [30]:

    
Y_test[0]









    Out[30]:





array([0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
       0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0])



In [31]:

    
f1_score(y_pred=Y_pred, y_true=Y_test, average='macro')









    



/home/hoangnt/anaconda3/lib/python3.5/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[31]:





0.14045250348537816



In [32]:

    
f1_score(y_pred=Y_pred, y_true=Y_test, average='micro')









    Out[32]:





0.29844588089671298



In [33]:

    
f1_score(y_pred=Y_pred, y_true=Y_test, average='macro')









    



/home/hoangnt/anaconda3/lib/python3.5/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[33]:





0.14045250348537816



In [ ]: