change source code and update jupyter notebook but retain variables

Question

I'm using a Jupyter notebook (Python2 kernel) to display the results of simulations that depend on a local library mylib that I've written. In the notebook, I have

[1]  ## import library
     import mylib

[2]  ## run simulations
     X = mylib.simulateX()
     Y = mylib.simulateY()
     Z = mylib.simulateZ()

[3]  ## plot data
     mylib.disp_data([X,Y,Z])

I often find myself making changes to one of the methods, say simulateY, but wanting to retain the results of existing simulations X and Z (or even just changing disp_data). That is, I'd like to reimport my library after changing the source code of simulateY to generate newY but still have access to X and Z. Is there a way to do this without writing X and Z to a file?

Possible duplicate of [How do I unload (reload) a Python module?](https://stackoverflow.com/questions/437589/how-do-i-unload-reload-a-python-module) — sal, Aug 22 '17 at 01:58

score 2 · Accepted Answer · answered Aug 22 '17 at 01:58

2

So I think that this might autoreload magic might work for you;

import my_lib
# Use these commands in the same cell.
%load_ext autoreload
%autoreload 2

answered Aug 22 '17 at 01:58

James Draper

5,110
5
40
59

change source code and update jupyter notebook but retain variables

1 Answers1