fix "except grammar" for python3, fix "pickle import" for python3, fi…

…x "list reference in class scope" for python3

datasets/Avazu.py

@@ -1,6 +1,6 @@
 import os
 
-from Dataset import Dataset
+from .Dataset import Dataset
 
 
 class Avazu(Dataset):
@@ -24,7 +24,8 @@ class Avazu(Dataset):
                   'C17', 'C18', 'C19', 'C20', 'C21', 'mday', 'hour', 'wday']
     feat_sizes = [7, 7, 3135, 3487, 24, 4002, 252, 28, 101449, 523672, 5925, 5, 4, 2417, 8, 9, 426, 4, 67, 166, 60, 10,
                   24, 7]
-    feat_min = [sum(feat_sizes[:i]) for i in range(max_length)]
+    feat_min = [0, 7, 14, 3149, 6636, 6660, 10662, 10914, 10942, 112391, 636063, 641988, 641993, 641997, 644414,
+                644422, 644431, 644857, 644861, 644928, 645094, 645154, 645164, 645188]
     data_dir = os.path.join(os.path.dirname(os.path.dirname(os.path.abspath(__file__))), 'Avazu')
     raw_data_dir = os.path.join(data_dir, 'raw')
     feature_data_dir = os.path.join(data_dir, 'feature')

datasets/Criteo.py

@@ -1,7 +1,7 @@
 import json
 import os
 
-from Dataset import Dataset
+from .Dataset import Dataset
 
 
 class Criteo(Dataset):
@@ -25,7 +25,9 @@ class Criteo(Dataset):
     feat_sizes = [4389, 8000, 329, 7432, 2646, 428, 233, 6301, 295, 11, 173, 176642,
                   585, 147117, 19845, 14830, 6916, 18687, 4, 6646, 1272, 46, 141085, 64381,
                   63692, 11, 2156, 7806, 61, 5, 928, 15, 147387, 116331, 145634, 57186, 9307, 63, 34]
-    feat_min = [sum(feat_sizes[:i]) for i in range(max_length)]
+    feat_min = [0, 4389, 12389, 12718, 20150, 22796, 23224, 23457, 29758, 30053, 30064, 30237, 206879, 207464, 354581,
+                374426, 389256, 396172, 414859, 414863, 421509, 422781, 422827, 563912, 628293, 691985, 691996, 694152,
+                701958, 702019, 702024, 702952, 702967, 850354, 966685, 1112319, 1169505, 1178812, 1178875]
     data_dir = os.path.join(os.path.dirname(os.path.dirname(os.path.abspath(__file__))), 'Criteo-8d')
     raw_data_dir = os.path.join(data_dir, 'raw')
     feature_data_dir = os.path.join(data_dir, 'feature')

datasets/Criteo_Challenge.py

@@ -1,10 +1,14 @@
 from __future__ import division
 from __future__ import print_function
 
-import cPickle as pkl
+import sys
+if sys.version.startswith('2'):
+    import cPickle as pkl
+else:
+    import pickle as pkl
 import os
 
-from Dataset import Dataset
+from .Dataset import Dataset
 
 
 class Criteo_Challenge(Dataset):

datasets/Criteo_all.py

@@ -1,13 +1,17 @@
 from __future__ import division
 from __future__ import print_function
 
-import cPickle as pkl
+import sys
+if sys.version.startswith('2'):
+    import cPickle as pkl
+else:
+    import pickle as pkl
 import os
 
 import numpy as np
 import pandas as pd
 
-from Dataset import Dataset, DatasetHelper
+from .Dataset import Dataset, DatasetHelper
 
 
 class Criteo_all(Dataset):

datasets/Dataset.py

@@ -294,7 +294,7 @@ def _iter_():
                         if squeeze_output:
                             y = y.squeeze()
                         yield X, y
-                    except StopIteration, e:
+                    except StopIteration:
                         print('finish', block)
                         break
             else:

datasets/__init__.py

@@ -1,11 +1,9 @@
 from .Criteo import Criteo
 from .iPinYou import iPinYou
 from .Avazu import Avazu
-from .Huawei import Huawei
-# from .Avazu_temp import Avazu_temp
-# from .iPinYousyf import iPinYousyf
-from Criteo_all import Criteo_all
-from Criteo_Challenge import Criteo_Challenge
+# from .Huawei import Huawei
+from .Criteo_all import Criteo_all
+from .Criteo_Challenge import Criteo_Challenge
 
 
 def as_dataset(data_name, initialized=True):
@@ -16,20 +14,11 @@ def as_dataset(data_name, initialized=True):
         return iPinYou(initialized=initialized)
     elif data_name == 'avazu':
         return Avazu(initialized=initialized)
-    elif data_name == 'huawei':
-        return Huawei(initialized=initialized)
     elif data_name == 'criteo_9d':
         return Criteo_all(initialized=initialized, num_of_days=9)
     elif data_name == 'criteo_16d':
         return Criteo_all(initialized=initialized, num_of_days=16)
     elif data_name == 'criteo_challenge':
         return Criteo_Challenge(initialized=initialized)
-
-
-# def as_temp(data_name, **kwargs):
-#     data_name = data_name.lower()
-#     if data_name == 'avazu':
-#         return Avazu_temp(**kwargs)
-#
-#     if data_name == 'ipinyou':
-#         return iPinYousyf(**kwargs)
+    # elif data_name == 'huawei':
+    #     return Huawei(initialized=initialized)

datasets/iPinYou.py

@@ -1,6 +1,6 @@
 import os
 
-from Dataset import Dataset
+from .Dataset import Dataset
 
 
 class iPinYou(Dataset):
@@ -26,7 +26,7 @@ class iPinYou(Dataset):
                 937623, 937664]
     feat_max = [7, 32, 704997, 705033, 705404, 705410, 756731, 937426, 937448, 937463, 937475, 937480, 937612, 937622,
                 937663, 937669]
-    feat_sizes = [feat_max[i] - feat_min[i] + 1 for i in range(max_length)]
+    feat_sizes = [8, 25, 704965, 36, 371, 6, 51321, 180695, 22, 15, 12, 5, 132, 10, 41, 6]
     data_dir = os.path.join(os.path.dirname(os.path.dirname(os.path.abspath(__file__))), 'iPinYou-all')
     raw_data_dir = os.path.join(data_dir, 'raw')
     feature_data_dir = os.path.join(data_dir, 'feature')

datasets/test.py

@@ -2,7 +2,7 @@
 
 import time
 
-from Criteo_Challenge import Criteo_Challenge
+from .Criteo_Challenge import Criteo_Challenge
 
 dataset = Criteo_Challenge(False)
 exit(0)