How to config nltk data directory from code?

Using fnjn's advice above on printing out the path:

print(nltk.data.path)

I saw the path strings in this type of format on windows:

C:\\Users\\my_user_name\\AppData\\Roaming\\SPB_Data

So I switched my path from the python type forward slash '/', to a double backslash '\\' when I used path.append:

nltk.data.path.append("C:\\workspace\\my_project\\data\\nltk_books")

The exception went away.

Another solution is to get ahead of it.

try import nltk nltk.download()

When the window box pops up asking if you want to download the corpus , you can specify there which directory it is to be downloaded to.