28 comments

[ 2.9 ms ] story [ 104 ms ] thread
I have prepared a few lessons on getting started with Python. It is basically a Python Textbook for those wanting to start learning Python for Data Science. In a unique twist, the textbook is a Jupyter Notebook in itself. Code and explanation all in one. You can directly run the code while you read the textbook, and master your Python skills. Currently includes Hello World, Language Semantics, Tuples and Lists. More lessons coming up.
I just started an ML course, and while I've been a developer for about a decade I've never used Python professionally. This will be a great help!
Awesome. Great to know. This is really encouraging.
I like the concept of Jupyter Notebooks and I do use them extensively, but I can't use my emacs shortcuts which makes writing anything more than a few lines really annoying.

Especially, writing Latex Math without Auctex (emacs package) is very tedious. What's the solution?

Have you tried to see if the shortcuts work on the Jupyter Labs version of the notebooks?
They interfere with the browser shortcuts, like Ctrl-N. Besides, still no auctex or similar.

I was literally thinking of making a custom chromium browser with my own shortcuts to solve this problem.

I have tried it. It works. But there is too much syntax surrounding code blocks that is annoying. And I need to enable inline Math display and inline image display to get a seamless interface.
Jupyter markdown blocks can parse LaTeX math and image embeddings. Pretty sure that's out of the box.
It is emacs, it is easy to eliminate the boiler-plate e.g., by using global/per file/per org sub tree defaults, templates.
I usually go the other route: instead of trying to get Jupyter notebooks to behave like my favorite editor, I try to get my favorite editor to behave like a Jupyter notebook. In vscode, the python extension can convert back and forth between python files and Jupyter notes books, and can run python files like Jupyter notebooks (with cells).
Do you know of any service offering hosted VSCode? The problem happens when my data is too large for my computer. In such cases it is much easier to use hosted Jupyter than run Python code on a server.
Have you tried remote development with VSCode? IT works really well! https://code.visualstudio.com/docs/remote/ssh
+1. Once you get it up and running it works so seamlessly.
I did try this, but it kind of did not work well for me. Since it requires a continuous ssh connection, and my connection kept dropping, the experience was not great. But maybe that was my internet. Need to give it a second attempt.
@moderators: This should be a show HN.
See also "Python Data Science Handbook" by Jake VanderPlas, which uses Jupyter notebooks too.

[0] https://jakevdp.github.io/PythonDataScienceHandbook/

I really like this, but I find it too exhaustive. It is too much to read. Wanted to create something that is short and to the point for programmers of other languages to quickly learn basics of Python for Data Science.
Have you seen Whirlwind Tour of Python by the same author? - https://jakevdp.github.io/WhirlwindTourOfPython/

> A Whirlwind Tour of Python is a fast-paced introduction to essential features of the Python language, aimed at researchers and developers who are already familiar with programming in another language. The material is particularly designed for those who wish to use Python for data science and/or scientific programming

Jupyter Book seems to be a good fit for this.
Yes I just saw. They launched it today itself. What a coincidence. If someone is familiar with it, please help migrate to Jupyter Book. Would appreciate a PR. Else will do it over the weekend.
what's the real difference between that vs Jupyter notebook vs Jupyter labs?
Jupyter Labs is an extension to Jupyter Notebook. Functionally the same, but Labs offers a much better coding interface. In it's simplest form, Jupyter Labs feels like a proper IDE, while just Jupyter feels like a pretty notebook offering an interface worthy of reading, but not worthy enough to used as a full time coding environment.
Is the mention of data science in the readme an indicator of future work that you want to add? Because as far as I can see there's no data science anywhere in the project so far.