3 ms·
I like the simplicity. I definitely get the payoff for standalone Python scripts, where once the script errors out the memory is cleared. But do you see a simi
by kapilsinha 3y ago
I like the simplicity. I definitely get the payoff for standalone Python scripts, where once the script errors out the memory is cleared.
But do you see a similar payoff for Jupyter notebooks (or similar)?
- williamzeng0 3y agoI think the marginal gain would be a lot less for Jupyter notebooks, but I've definitely rerun individual cells and wasted time there before. I think it could help if you forget to save the output of a function within a single cell like this: 1. print(f(x)) # -> check what happened 2. out = f(x) # -> turns out we want to save this, so we have to wait again
- p10_user 3y agoFWIW There is a built-in cache system for r markdown documents. I'm not up to speed on their exact implementation but I have found it useful. https://bookdown.org/yihui/rmarkdown-cookbook/cache.html https://bookdown.org/yihui/rmarkdown-cookbook/cache.html
- williamzeng0 3y agofile.mtime. (file last modified) is an awesome way to key the cache.