Explanation for output vector . What are they representing and how to decode them .
like : for "Fun camera, not a fun amazon experience"
output is " 0 1 0 0 1 0 0 1 0 0 1 0 0 1 0 "
what are these "0"s and "1"s are representing .

请问下获取京东评论里的找不到输入输出路径呢，那个数据集没有，在哪儿找呢

The line 6 has to be changed from from blstm_sw import BLSTM to from blstm_sw import BLSTM. It is just a simple import error.

After running
python RunRNN.py
for a long while, at the end, I get this error:

  [33600]: mean loss 0.121167
  Seen 34000 in 1886.55 s
  [34000]: mean loss 0.137862
  Seen 34400 in 1909.40 s
  [34400]: mean loss 0.1298
  Seen 34800 in 1932.91 s
  [34800]: mean loss 0.120234
  [35000]: mean loss 0.121845
SGD complete: 35000 examples in 1948.43 seconds.
Traceback (most recent call last):
  File "RunRNN.py", line 85, in <module>
    score1.append(build_confusion_matrix(X_dev,Y_dev,model))
  File "/scratch2/debate_tweets/sentiment/EntitySentiment/misc.py", line 139, in build_confusion_matrix
    conf_arr[true_label,guessed_label]+=1
IndexError: only integers, slices (`:`), ellipsis (`...`), numpy.newaxis (`None`) and integer or boolean arrays are valid indices

ValueError Traceback (most recent call last)
in ()
10 model = BRNN(L0, U0=None, alpha=0.05, rseed=10, bptt=20)
11 #print w
---> 12 curve = model.train_sgd(X,Y,idxiter_random,None,400,400)
13 score1.append(build_confusion_matrix(X_dev,Y_dev,model))
14 print score1

/Users/Desktop/Python/EntitySentiment-master/nn/base.py in train_sgd(self, X, y, idxiter, alphaiter, printevery, costevery, devidx)
384 printevery=10000, costevery=10000,
385 devidx=None):
--> 386 if idxiter==None: # default training schedule
387 idxiter = xrange(len(y))
388 if alphaiter == None: # default training schedule

ValueError: The truth value of an array with more than one element is ambiguous. Use a.any() or a.all()

The merge conflicts in misc.py have not been removed. Please clean the code.