The problem recurred

When I use pickle to reload data, all the errors are as follows:

Traceback (most recent call last):
  File "segment.py", line 17, in <module>
    word2id = pickle.load(pk)
  File "/System/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/pickle.py", line 1378, in load
    return Unpickler(file).load()
  File "/System/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/pickle.py", line 858, in load
    dispatch[key](self)
  File "/System/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/pickle.py", line 1090, in load_global
    klass = self.find_class(module, name)
  File "/System/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/pickle.py", line 1124, in find_class
    __import__(module)
ImportError: No module named indexes.base

The reason for this

The same code and data run on two different machines. At first, I thought the wrong machine was missing some Python packages. But there are too many packages to install, so I can’t try them one by one. Fortunately, I use virsualenv to copy the environment from another machine to this machine directly. After running, there is no problem. But in order to find out which Python installation package is missing, I use the original compilation environment, reuse pickle to generate the original data to be loaded, and then reload it At this time, there was no error.

summary

To sum up, the reason is that the original version of panda used in the generation of pickle file is different from the current version of load pickle file. So whether it is to write code in Python or other languages, the compiling environment is very important. Once the version of a package is different, it may also lead to program errors.