AttributeError: module 'preprocessor' has no attribute 'clean'

user_2965031 0
Viewed 4288

I am trying to use the preprocessor library in order to clean text stored in a Pandas Data Frame. I've installed the last version (https://pypi.org/project/tweet-preprocessor/), but I receive this error message:

import preprocessor as p
#forming a separate feature for cleaned tweets
for i,v in enumerate(df['text']):
    df.loc[v,'text'] = p.clean(i)

---------------------------------------------------------------------------
AttributeError                            Traceback (most recent call last)
<ipython-input-183-94e08e1aff33> in <module>
      1 #forming a separate feature for cleaned tweets
      2 for i,v in enumerate(df['text']):
----> 3     df.loc[v,'text'] = p.clean(i)

AttributeError: module 'preprocessor' has no attribute 'clean'
attributeerror preprocessor python
2 Answers

You probably have the preprocessor module installed as well, which is entirely distinct from the tweet-preprocessor module. However, confusingly, the import preprocessor as p statement can be used for both. When both modules are installed, Python ignores tweet-preprocessor and automatically opts for preprocessor, which does not contain a clean function, hence the error you received.

To resolve this, I had to uninstall both modules with the following commands:

pip uninstall preprocessor
pip uninstall tweet-preprocessor

Then I closed all shells for a fresh start and typed:

pip install tweet-preprocessor

And finally:

>>> import preprocessor as p
>>> p.clean('#this and that')
'and that'

Merely uninstalling preprocessor did not work. Python kept importing the module despite it being uninstalled. I am not sure why, but I suspect it has something to do with caches that Python keeps in the background.

user_156112170

Try installing first:

pip install tweet-preprocessor

Then:

import preprocessor as p
user_141371750
Related