3/30/2023

First, install mamba as a lighter weight and faster alternative to conda
Next, install jupyter notebook server which relies on mamba + python
Also read this article on Managing Python Environment in 2022
Less necessary, but read this section of this longer, more comprehensive article on understanding venv in particular. But it also has comments on virtualenv, conda, etc.
Ross Dobson’s January 2022 tutorial on setting up Python environments with Mambaforge
This 2020 update to an older 2018 version of a post explains why “Pip is here to stay!”
Older 2018 piece on “Why You Need Python Environments and How to Manage Them with Conda”
Cached version of this older Coiled blog post on “Use Mambaforge to Conda Install PyData Stack on your Apple M1 Silicon Machine”

4/01/2023

Quote from aseifert’s article: “Pip, venv, virtualenv, pyenv, pipenv, micropipenv, pip-tools, conda, miniconda, mamba, micromamba, poetry, hatch, pdm, pyflow 🤯 These days even the most senior Python developer is confused about all the options to manage environments.”
After successful install of mambaforge from github (see ON notes), i verified that both conda info and mamba info work. Eventually, I should read this page from the mamba docs site.
Took multiple steps, but using this YouTube videoI was able to: (1) install mamba, (2) install python 3.10 within mamba, (3) install jupyter lab, (4) activate the jupyter mamba ‘space’ or whatever it’s called, and (5) launch the jupyter server and connect to it from my local web client.
Out of curiosity, I wondered how well PyTorch works with Apple Silicon and was led by this reddit comment to this PyTorch GitHub issue
Should go through the basic PyTorch tutorial and download the Jupyter notebook to run locally or can run through Google Colab.

Mamba + Jupyter setup video How to install JupyterLab using Mamba
In the proper mamba environment, run ‘jupyter notebook’ and it should start the server in the terminal and automatically open a web browser client. Success!

Installing pytorch using conda / mamba. Gated Medium Towards Data Science article
Successfully installed stable 2.0.0 version for Apple Silicon using ‘mamba install pytorch torchvision -cpytorch’

To see list of conda environments created, use this command ‘conda info –envs’
Steps:
1. install conda by navigating to mamba-forge at github. Download the “Mambaforge-MacOSX-arm64.sh”, navigate to the directory, and then execute the shell script by typing ‘Mambaforge-MacOSX-arm64.sh’ from the command line.
2. create an environment for pytorch. E.g., I want to call my personal pytorch environment pytorch-jh, so i type ‘mamba create -n pytorch-jh -c conda-forge python=3.10’
  - ‘create -n pytorch-jh’ creates the environment called pytorch-jh
  - ‘create -c conda-forge’ indicates i need to download python from the conda-forge channel
  - ‘python=3.10’ pins the version of python as 3.10 for the pytorch-jh environment
3. Download jupyter packages into pytorch-jh with this command ‘mamba install -n pytorch-jh -c conda-forge jupyterlab’
4. Download PyTorch packages into pytorch-jh with this command ‘mamba install -n pytorch-jh pytorch torchvision -cpytorch’
  - ‘install -n pytorch-jh’ means install into the pytorch-jh environmnet
  - the list of packages per PyTorch.org are: pytorch, torchvision,
  - i’m not exactly sure what ‘-cpytorch’ is doing but it somehow knew to download the mac-os-arm64 version so I think this is related
Verified that pytorch is working and using the M2 GPU by running 1jh_test.py and 2jh_test.py adapted from the above Towards Data Science article

Used ChatGPT to find latest in Python books and courses
Thoughts and notes on lists (ordered) vs. dictionaries (unorderd key-value pairs, allowing duplicate values) vs. sets (like dicts but unique with no distinct values)
examples of iterables. All of the 3 above plus strings. See
zip() function. strip() / lstrip() / rstrip() to get rid of whitespace characters. difference between pop() and delete().
using python interpreter, up to ‘Organizing a List’ on p. 42
when in doubt about using an array versus a standard python list, use the better, more mathematically useful numpy.array()
- Do not default to array.array() which is a C-style array. Instead, use the more mathematically useful numpy.array()

Simple explanation for ‘+=’ operator. Initialize ‘a = 10’. To set a to 17, you can then type ‘a += 7’. Then, print(a) will return ‘17’.

Found via phind.com’s answer about tensors vs. multi-dimensional array. This article a pretty good simple way of getting familiar with PyTorch tensors. Requires import of torch, numpy, and pandas

Comprehensions are useful syntatic sugar for quickly populating and looping through both lists and dictionaries.
Concept of “collections” useful noun as explained in this YouTube about list comprehensions (80 seconds in)

Bunch of info on Python environment/version/package managers on Hacker News today.
- This article by Larry Du summarizes state of the art as of July 2024. HN thread.
- Introduction to Rye, Intro YouTube video, and HN thread.
- In 2024, the core developers behind mamba have also started working on uv a new Python packager written in Rust. Per the Larry Du article above, “uv is by far the most promising package management tool in the Python ecosystem as of the writing of this post. This project actually aims to be a drop-in replacement for pip on top of being a Cargo for Python. The API is currently in no ways stable (as of 2024), but the benchmarks are incredibly promsing. Most notably, the development is backed by Astral.sh, a company formed by Charlie Marsh and the creators of the ruff linter, a widely beloved tool that virtually supplanted all incumbants overnight when it was released in 2022.”