How to refresh a Python import in a Jupyter Notebook cell?

Question

I have two files:

MyModule.py MyNotebook.ipynb

I am using Jupyter Notebook, latest, and Python, latest. I have two code cells in my Notebook.

Cell #1

import some stuff
Run some code
(Keep everything in the environment takes about five minutes to run this cell).

Cell #2

import MyModule
Execute code from MyModule

I would like to make code changes in MyModule.py and rerun Cell #2 but without restarting the kernel ( Cell #1 did a fair amount of work which I don't want to rerun each time). If I simply run the second cell, those changes made to MyModule.py do not propagate through. I did some digging, and I tried using importlib.reload. The actual code for Cell #2 :

from Nash.IOEngineNash import *
import importlib
importlib.reload(Nash.IOEngineNash)

Unfortunately, this isn't quite working. How can I push those changes in MyModule.py (or Nash/IOEngineNash.py in actual fact) into my Notebook without restarting the kernel and running from scratch?

Answer 1

I faced a similar issue, while importing a custom script in jupyter notebook

Try importing the module as an alias then reloading it

import Nash as nash
from importlib import reload
reload(nash)

Answer 2

To make sure that all references to the old version of the module are updated, you might want to re-import it after reloading, eg

import mymodule
reload(mymodule)
import mymodule

Issues may arise due to implicit dependencies, because importlib.reload only reloads the (top-level) module that you ask it to reload, not its submodules or external modules.

If you want to reload a module recursively, you might have a look at this gist , which allows you to simply add a line like this at the top of a notebook cell:

%reload mymodule
import mymodule

This recursively reloads the mymodule and all of its submodules.