magma-nvim-goose VS nvim-ipython-repl

magma-nvim for interactive evaluating

Not with notebooks, but for interactive evaluation I use my fork of magma (gotta rename one day but lazy) + language servers (whichever you like, I use jedi) + jupyter kernel for auto completion. Why I like it more than notebooks

I use my fork of magma and trying out right now jupyter-kernel, which hopefully will provide autocompletion.

Interesting. I use magma-nvim (to be precise, my fork, bc the upstream seems to be dead) for sending code and interactive evaluation from jupyter kernels. I'd like autocompletion for sure (for now, I use LSP, but it doesn't work too well, because you know, it analyzes the file from up down, not like in jupyter where I can run any arbitrary code)

Not a solution for vim, but for nvim I use my fork of magma https://github.com/WhiteBlackGoose/magma-nvim-goose

I use my fork of magma-nvim. To view graphs from matplotlib and the like, I just save them to a file and run nsxiv on the produced pic. (but magma also supports showing graphs with ueberzueg, assuming you have a terminal that supports it)

You use magma-nvim for interactive jupyter experience (or my fork which also includes three last PRs)

I ended up writing my own Python one because I wanted to better control what happens when sending bigger code blocks to the repl. I'd generally steer folks to larger solutions, but if you are just using python and want to just see a snippet of what you sent echoed in the repl, then this could fit your need: https://github.com/shea-parkes/nvim-ipython-repl

My send-to-repl (very rough around the edges, sharing just as an FYI): https://github.com/shea-parkes/nvim-ipython-repl