In [1]:

    

df = pd.read_csv('../resource/final_df2.csv')
y = df['rating(y)'].values
real_X = df.ix[:,'avg_rating':'skew']
real_X.head(2)


    

    
Out[1]:






  

    

      

      
avg_rating
      
lee_rating
      
eval_count
      
wish_count
      
cmt_count
      
run_time
      
year
      
std
      
skew
    
  
  

    

      
0
      
4.22683
      
4
      
13025
      
9796
      
2585
      
128
      
2015
      
0.625521
      
-1.002260
    
    

      
1
      
2.99629
      
3
      
58122
      
3166
      
965
      
121
      
2013
      
1.059023
      
-0.347552
    
  

In [2]:

    

from sklearn.cross_validation import KFold, StratifiedKFold
from sklearn.naive_bayes import GaussianNB
from sklearn.metrics import mean_absolute_error
import itertools

gnb = GaussianNB()
cv = StratifiedKFold(y, n_folds=3, random_state=0)


    

In [3]:

    

scores = np.zeros(3)
for i, (train_idx, test_idx) in enumerate(cv):
    X_train = real_X.ix[train_idx]
    y_train = y[train_idx]
    X_test = real_X.ix[test_idx]
    y_test = y[test_idx]
    
    gnb.fit(X_train, y_train)
    y_pred = gnb.predict(X_test)
    
    scores[i] = mean_absolute_error(y_test, y_pred)

scores
print('mean:', round(np.mean(scores), 4), '   ', scores)


    

    




mean: 1.0493     [ 1.03825137  1.12087912  0.98882682]

In [4]:

    

combi_df1 = pd.DataFrame(columns=['combi', 'mean', 'scores'])

for c in range(9):
    new_X = real_X[[c]]
    
    scores = np.zeros(3)
    for i, (train_idx, test_idx) in enumerate(cv):
        X_train = new_X.ix[train_idx]
        y_train = y[train_idx]
        X_test = new_X.ix[test_idx]
        y_test = y[test_idx]
    
        gnb.fit(X_train, y_train)
        y_pred = gnb.predict(X_test)
    
        scores[i] = mean_absolute_error(y_test, y_pred)

    scores
    combi_df1.loc[len(combi_df1)] = [c, round(np.mean(scores), 4), scores]
    #print(c, 'mean:', round(np.mean(scores), 4), '   ', scores)


    

In [5]:

    

print(len(combi_df1))
combi_df1.sort_values('mean', ascending=True).head(5)


    

    




9






    
Out[5]:






  

    

      

      
combi
      
mean
      
scores
    
  
  

    

      
0
      
0
      
0.6374
      
[0.655737704918, 0.692307692308, 0.564245810056]
    
    

      
8
      
8
      
0.6414
      
[0.633879781421, 0.67032967033, 0.620111731844]
    
    

      
7
      
7
      
0.7570
      
[0.79781420765, 0.758241758242, 0.715083798883]
    
    

      
5
      
5
      
0.7867
      
[0.814207650273, 0.763736263736, 0.782122905028]
    
    

      
6
      
6
      
0.8310
      
[0.825136612022, 0.818681318681, 0.849162011173]
    
  

In [6]:

    

combi_df2 = pd.DataFrame(columns=['combi', 'mean', 'scores'])

idx_list = list(itertools.combinations(range(9), 2))
for idx in idx_list:
    new_X = real_X[list(idx)]
    
    scores = np.zeros(3)
    for i, (train_idx, test_idx) in enumerate(cv):
        X_train = new_X.ix[train_idx]
        y_train = y[train_idx]
        X_test = new_X.ix[test_idx]
        y_test = y[test_idx]
    
        gnb.fit(X_train, y_train)
        y_pred = gnb.predict(X_test)
    
        scores[i] = mean_absolute_error(y_test, y_pred)

    scores
    combi_df2.loc[len(combi_df2)] = [idx, round(np.mean(scores), 4), scores]
    #print(idx, 'mean:', round(np.mean(scores), 4), '   ', scores)


    

In [7]:

    

print(len(combi_df2))
combi_df2.sort_values('mean', ascending=True).head(5)


    

    




36






    
Out[7]:






  

    

      

      
combi
      
mean
      
scores
    
  
  

    

      
4
      
(0, 5)
      
0.6541
      
[0.661202185792, 0.708791208791, 0.59217877095]
    
    

      
5
      
(0, 6)
      
0.6559
      
[0.661202185792, 0.708791208791, 0.597765363128]
    
    

      
32
      
(5, 8)
      
0.6762
      
[0.68306010929, 0.708791208791, 0.63687150838]
    
    

      
34
      
(6, 8)
      
0.6912
      
[0.693989071038, 0.686813186813, 0.692737430168]
    
    

      
0
      
(0, 1)
      
0.6913
      
[0.628415300546, 0.752747252747, 0.692737430168]
    
  

In [8]:

    

combi_df3 = pd.DataFrame(columns=['combi', 'mean', 'scores'])

idx_list = list(itertools.combinations(range(9), 3))
for idx in idx_list:
    new_X = real_X[list(idx)]
    
    scores = np.zeros(3)
    for i, (train_idx, test_idx) in enumerate(cv):
        X_train = new_X.ix[train_idx]
        y_train = y[train_idx]
        X_test = new_X.ix[test_idx]
        y_test = y[test_idx]
    
        gnb.fit(X_train, y_train)
        y_pred = gnb.predict(X_test)
    
        scores[i] = mean_absolute_error(y_test, y_pred)

    scores
    combi_df3.loc[len(combi_df3)] = [idx, round(np.mean(scores), 4), scores]
    #print(idx, 'mean:', round(np.mean(scores), 4), '   ', scores)


    

In [9]:

    

print(len(combi_df3))
combi_df3.sort_values('mean').head(5)


    

    




84






    
Out[9]:






  

    

      

      
combi
      
mean
      
scores
    
  
  

    

      
22
      
(0, 5, 6)
      
0.6633
      
[0.666666666667, 0.708791208791, 0.614525139665]
    
    

      
26
      
(0, 6, 8)
      
0.6910
      
[0.688524590164, 0.71978021978, 0.664804469274]
    
    

      
3
      
(0, 1, 5)
      
0.6928
      
[0.677595628415, 0.758241758242, 0.642458100559]
    
    

      
82
      
(5, 7, 8)
      
0.6967
      
[0.710382513661, 0.675824175824, 0.703910614525]
    
    

      
4
      
(0, 1, 6)
      
0.6970
      
[0.655737704918, 0.703296703297, 0.731843575419]
    
  

In [10]:

    

combi_df4 = pd.DataFrame(columns=['combi', 'mean', 'scores'])

idx_list = list(itertools.combinations(range(9), 4))
for idx in idx_list:
    new_X = real_X[list(idx)]
    
    scores = np.zeros(3)
    for i, (train_idx, test_idx) in enumerate(cv):
        X_train = new_X.ix[train_idx]
        y_train = y[train_idx]
        X_test = new_X.ix[test_idx]
        y_test = y[test_idx]
    
        gnb.fit(X_train, y_train)
        y_pred = gnb.predict(X_test)
    
        scores[i] = mean_absolute_error(y_test, y_pred)

    scores
    combi_df4.loc[len(combi_df4)] = [idx, round(np.mean(scores), 4), scores]


    

In [11]:

    

print(len(combi_df4))
combi_df4.sort_values('mean').head(5)


    

    




126






    
Out[11]:






  

    

      

      
combi
      
mean
      
scores
    
  
  

    

      
15
      
(0, 1, 5, 6)
      
0.7040
      
[0.672131147541, 0.758241758242, 0.68156424581]
    
    

      
16
      
(0, 1, 5, 7)
      
0.7111
      
[0.72131147541, 0.747252747253, 0.664804469274]
    
    

      
55
      
(0, 6, 7, 8)
      
0.7130
      
[0.732240437158, 0.730769230769, 0.675977653631]
    
    

      
53
      
(0, 5, 6, 8)
      
0.7132
      
[0.699453551913, 0.736263736264, 0.703910614525]
    
    

      
54
      
(0, 5, 7, 8)
      
0.7188
      
[0.737704918033, 0.686813186813, 0.731843575419]
    
  

In [12]:

    

combi_df5 = pd.DataFrame(columns=['combi', 'mean', 'scores'])

idx_list = list(itertools.combinations(range(9), 5))
for idx in idx_list:
    new_X = real_X[list(idx)]
    
    scores = np.zeros(3)
    for i, (train_idx, test_idx) in enumerate(cv):
        X_train = new_X.ix[train_idx]
        y_train = y[train_idx]
        X_test = new_X.ix[test_idx]
        y_test = y[test_idx]
    
        gnb.fit(X_train, y_train)
        y_pred = gnb.predict(X_test)
    
        scores[i] = mean_absolute_error(y_test, y_pred)

    scores
    combi_df5.loc[len(combi_df5)] = [idx, round(np.mean(scores), 4), scores]


    

In [13]:

    

print(len(combi_df5))
combi_df5.sort_values('mean').head(5)


    

    




126






    
Out[13]:






  

    

      

      
combi
      
mean
      
scores
    
  
  

    

      
32
      
(0, 1, 5, 6, 8)
      
0.7098
      
[0.655737704918, 0.730769230769, 0.743016759777]
    
    

      
69
      
(0, 5, 6, 7, 8)
      
0.7297
      
[0.743169398907, 0.725274725275, 0.720670391061]
    
    

      
31
      
(0, 1, 5, 6, 7)
      
0.7298
      
[0.68306010929, 0.796703296703, 0.709497206704]
    
    

      
33
      
(0, 1, 5, 7, 8)
      
0.7300
      
[0.704918032787, 0.730769230769, 0.754189944134]
    
    

      
104
      
(1, 5, 6, 7, 8)
      
0.7317
      
[0.688524590164, 0.769230769231, 0.737430167598]
    
  

In [14]:

    

combi_df6 = pd.DataFrame(columns=['combi', 'mean', 'scores'])

idx_list = list(itertools.combinations(range(9), 6))
for idx in idx_list:
    new_X = real_X[list(idx)]
    
    scores = np.zeros(3)
    for i, (train_idx, test_idx) in enumerate(cv):
        X_train = new_X.ix[train_idx]
        y_train = y[train_idx]
        X_test = new_X.ix[test_idx]
        y_test = y[test_idx]
    
        gnb.fit(X_train, y_train)
        y_pred = gnb.predict(X_test)
    
        scores[i] = mean_absolute_error(y_test, y_pred)

    scores
    combi_df6.loc[len(combi_df6)] = [idx, round(np.mean(scores), 4), scores]


    

In [15]:

    

print(len(combi_df6))
combi_df6.sort_values('mean').head(5)


    

    




84






    
Out[15]:






  

    

      

      
combi
      
mean
      
scores
    
  
  

    

      
54
      
(0, 3, 5, 6, 7, 8)
      
0.7443
      
[0.737704918033, 0.78021978022, 0.715083798883]
    
    

      
34
      
(0, 1, 5, 6, 7, 8)
      
0.7483
      
[0.699453551913, 0.796703296703, 0.748603351955]
    
    

      
28
      
(0, 1, 3, 5, 7, 8)
      
0.7496
      
[0.79781420765, 0.758241758242, 0.692737430168]
    
    

      
27
      
(0, 1, 3, 5, 6, 8)
      
0.7591
      
[0.726775956284, 0.818681318681, 0.731843575419]
    
    

      
29
      
(0, 1, 3, 6, 7, 8)
      
0.7702
      
[0.732240437158, 0.824175824176, 0.754189944134]
    
  

In [16]:

    

combi_df7 = pd.DataFrame(columns=['combi', 'mean', 'scores'])

idx_list = list(itertools.combinations(range(9), 7))
for idx in idx_list:
    new_X = real_X[list(idx)]
    
    scores = np.zeros(3)
    for i, (train_idx, test_idx) in enumerate(cv):
        X_train = new_X.ix[train_idx]
        y_train = y[train_idx]
        X_test = new_X.ix[test_idx]
        y_test = y[test_idx]
    
        gnb.fit(X_train, y_train)
        y_pred = gnb.predict(X_test)
    
        scores[i] = mean_absolute_error(y_test, y_pred)

    scores
    combi_df7.loc[len(combi_df7)] = [idx, round(np.mean(scores), 4), scores]


    

In [17]:

    

print(len(combi_df7))
combi_df7.sort_values('mean').head(5)


    

    




36






    
Out[17]:






  

    

      

      
combi
      
mean
      
scores
    
  
  

    

      
19
      
(0, 1, 3, 5, 6, 7, 8)
      
0.7591
      
[0.726775956284, 0.813186813187, 0.737430167598]
    
    

      
20
      
(0, 1, 4, 5, 6, 7, 8)
      
0.7883
      
[0.814207650273, 0.807692307692, 0.743016759777]
    
    

      
14
      
(0, 1, 2, 5, 6, 7, 8)
      
0.8255
      
[0.841530054645, 0.774725274725, 0.860335195531]
    
    

      
17
      
(0, 1, 3, 4, 5, 7, 8)
      
0.8635
      
[0.934426229508, 0.846153846154, 0.810055865922]
    
    

      
27
      
(0, 3, 4, 5, 6, 7, 8)
      
0.8678
      
[0.868852459016, 0.840659340659, 0.893854748603]
    
  

In [18]:

    

combi_df8 = pd.DataFrame(columns=['combi', 'mean', 'scores'])

idx_list = list(itertools.combinations(range(9), 8))
for idx in idx_list:
    new_X = real_X[list(idx)]
    
    scores = np.zeros(3)
    for i, (train_idx, test_idx) in enumerate(cv):
        X_train = new_X.ix[train_idx]
        y_train = y[train_idx]
        X_test = new_X.ix[test_idx]
        y_test = y[test_idx]
    
        gnb.fit(X_train, y_train)
        y_pred = gnb.predict(X_test)
    
        scores[i] = mean_absolute_error(y_test, y_pred)

    scores
    combi_df8.loc[len(combi_df8)] = [idx, round(np.mean(scores), 4), scores]


    

In [19]:

    

print(len(combi_df8))
combi_df8.sort_values('mean').head(5)


    

    




9






    
Out[19]:






  

    

      

      
combi
      
mean
      
scores
    
  
  

    

      
6
      
(0, 1, 3, 4, 5, 6, 7, 8)
      
0.8619
      
[0.852459016393, 0.912087912088, 0.821229050279]
    
    

      
4
      
(0, 1, 2, 3, 5, 6, 7, 8)
      
0.8840
      
[0.868852459016, 0.93956043956, 0.843575418994]
    
    

      
0
      
(0, 1, 2, 3, 4, 5, 6, 7)
      
1.0493
      
[1.03825136612, 1.12087912088, 0.988826815642]
    
    

      
1
      
(0, 1, 2, 3, 4, 5, 6, 8)
      
1.0493
      
[1.03825136612, 1.12087912088, 0.988826815642]
    
    

      
7
      
(0, 2, 3, 4, 5, 6, 7, 8)
      
1.0493
      
[1.03825136612, 1.12087912088, 0.988826815642]