ImportError: cannot import name 'ensure_dir_exists'


I update the Jupyter notebook from the old version to latest 5.3.1. However, when I try to launch the notebook from anaconda, it throws an import error:
I tried to remove and install Jupyter package, still, the issue persists.

Traceback (most recent call last):
File "", line 6, in 
from notebook.notebookapp import main
File "", line 25, in 
from .nbextensions import install_nbextension
File "", line 27, in 
from jupyter_core.utils import ensure_dir_exists
ImportError: cannot import name 'ensure_dir_exists'
Asked By: Santosh Kangane



You need to update jupyter_core and jupyter_client manually from your terminal:

conda update jupyter_core jupyter_client
Answered By: piratefache

If you’ve installed jupyter using pip3.

pip3 install --upgrade jupyter_core jupyter_client
Answered By: pavan

I’ve met the same problem when I use ipython notebook .

After I tried upgrade jupyter, then I tried reinstall the ipython.
The import error seem fixed.

pip3 uninstall ipython

pip3 install ipython

Answered By: Wu Wenter

update jupyter core

pip install jupyter-core --upgrade

this working for me in Anaconda3 with python 3.6

Answered By: Cristian Q

When I tried to install R the notebook was upgraded to 5.3.1 and same problem appeared, however the highest vote answer apparently didn’t work for me.

What I did was:

conda uninstall jupyter notebook


conda install jupyter notebook

and the problem solved with the old version.

Answered By: Pistachio Guoguo

I had a similar issue on Ubuntu after installing RStudio on my system. To solve it, first I tried @MichaelCara suggestion to update jupyter. However, after that I got the following new message error:

ImportError: cannot open shared object file: No such file or directory

which I solved installing libsodium as explained here.

In short:

conda update jupyter_core jupyter_client jupyter_console
conda install libsodium

Answered By: tbnsilveira
Categories: questions Tags: ,
Answers are sorted by their score. The answer accepted by the question owner as the best is marked with
at the top-right corner.