Moved cPickle to six cPickle

python-package-requirement.txt

@@ -12,6 +12,7 @@ numpy>=1.11
 pandas
 requests
 scipy>=0.18
+six
 sqlalchemy>=1.0.14
 tensorflow>=1.0
 tika

snorkel/learning/logistic_regression.py

@@ -1,10 +1,10 @@
-import six.moves.cPickle
 import numpy as np
 import tensorflow as tf
 
 from snorkel.learning.disc_learning import TFNoiseAwareModel
 from scipy.sparse import csr_matrix, issparse
 from time import time
+from six.moves.cPickle import dump, load
 from snorkel.learning.utils import LabelBalancer
 
 
@@ -137,11 +137,11 @@ def get_weights(self):
 
     def save_info(self, model_name):
         with open('{0}.info'.format(model_name), 'wb') as f:
-            cPickle.dump((self.d, self.lr, self.l1_penalty, self.l2_penalty), f)
+            dump((self.d, self.lr, self.l1_penalty, self.l2_penalty), f)
 
     def load_info(self, model_name):
         with open('{0}.info'.format(model_name), 'rb') as f:
-            self.d, self.lr, self.l1_penalty, self.l2_penalty = cPickle.load(f)
+            self.d, self.lr, self.l1_penalty, self.l2_penalty = load(f)
 
 
 class SparseLogisticRegression(LogisticRegression):

test/ParserTests.py

@@ -1,6 +1,6 @@
 import os, requests, sys, unittest
 sys.path.insert(1, os.path.join(sys.path[0], '..'))
-import cPickle
+from six.moves.cPickle import load
 from snorkel.parser import *
 
 ROOT = os.environ['SNORKELHOME']
@@ -20,10 +20,10 @@ def test_parser_1(self):
 
         # Load correct parses
         with open(ROOT + '/test/data/CDR_TestSet_docs.pkl', 'rb') as f:
-            gold_docs = cPickle.load(f)
+            gold_docs = load(f)
 
         with open(ROOT + '/test/data/CDR_TestSet_sents.pkl', 'rb') as f:
-            gold_sents = cPickle.load(f)
+            gold_sents = load(f)
 
         # Set up the doc parser
         xml_parser = XMLDocParser(

tutorials/cdr/CDR_Tutorial_1.ipynb

@@ -143,7 +143,7 @@
    },
    "outputs": [],
    "source": [
-    "import cPickle\n",
+    "import six.moves.cPickle\n",
     "\n",
     "with open('data/doc_ids.pkl', 'rb') as f:\n",
     "    train_ids, dev_ids, test_ids = cPickle.load(f)\n",

tutorials/cdr/CDR_Tutorial_2.ipynb

@@ -81,7 +81,7 @@
    "outputs": [],
    "source": [
     "import bz2\n",
-    "import cPickle\n",
+    "import six.moves.cPickle\n",
     "\n",
     "with bz2.BZ2File('data/ctd.pkl.bz2', 'rb') as ctd_f:\n",
     "    ctd_unspecified, ctd_therapy, ctd_marker = cPickle.load(ctd_f)"

tutorials/cdr/id_script.py

@@ -1,6 +1,6 @@
-import cPickle
 import numpy as np
 from collections import defaultdict
+from six.moves.cPickle import dump
 
 def cdr_doc_split(
     train_fname='data/CDR_TrainingSet.BioC.xml',
@@ -32,7 +32,7 @@ def cdr_doc_split(
     )))
     # Save to file
     with open(out_fname, 'wb') as f:
-        cPickle.dump((id_dict['train'], id_dict['dev'], id_dict['test']), f)
+        dump((id_dict['train'], id_dict['dev'], id_dict['test']), f)
 
 if __name__ == '__main__':
 	cdr_doc_split()

tutorials/cdr/load_external_annotations.py

@@ -1,4 +1,4 @@
-import cPickle
+from six.moves.cPickle import load
 
 from snorkel.db_helpers import reload_annotator_labels
 from snorkel.models import StableLabel
@@ -8,7 +8,7 @@ def load_external_labels(session, candidate_class, split, annotator='gold',
     label_fname='data/cdr_relations_gold.pkl', id_fname='data/doc_ids.pkl'):
     # Load document-level relation annotations
     with open(label_fname, 'rb') as f:
-        relations = cPickle.load(f)
+        relations = load(f)
     # Get split candidates
     candidates = session.query(candidate_class).filter(
         candidate_class.split == split

tutorials/cdr/utils.py

@@ -1,5 +1,5 @@
 import bz2
-import cPickle
+from six.moves.cPickle import load
 
 from string import punctuation
 
@@ -21,7 +21,7 @@ class CDRTagger(object):
 
     def __init__(self, fname='data/unary_tags.pkl.bz2'):   
         with bz2.BZ2File(fname, 'rb') as f:
-            self.tag_dict = cPickle.load(f)
+            self.tag_dict = load(f)
 
     def tag(self, parts):
         pubmed_id, _, _, sent_start, sent_end = parts['stable_id'].split(':')
@@ -44,9 +44,9 @@ class TaggerOneTagger(CDRTagger):
     def __init__(self, fname_tags='data/taggerone_unary_tags_cdr.pkl.bz2',
         fname_mesh='data/chem_dis_mesh_dicts.pkl.bz2'):
         with bz2.BZ2File(fname_tags, 'rb') as f:
-            self.tag_dict = cPickle.load(f)
+            self.tag_dict = load(f)
         with bz2.BZ2File(fname_mesh, 'rb') as f:
-            self.chem_mesh_dict, self.dis_mesh_dict = cPickle.load(f)
+            self.chem_mesh_dict, self.dis_mesh_dict = load(f)
 
     def tag(self, parts):
         parts = super(TaggerOneTagger, self).tag(parts)