Gå til hovedinnhold

Python

Scripting​

I find Python to be very useful for scripting. I use scripting for a lot of small things, such as working on text parsing, simple automation and others. I did, for example create a script for parsing book quotes from the Kindle (Here is the script)

Ecosystem​

Managing dependencies in Python is one of the most challenging and annoying things to work on. You should not only understand the differences between Conda, pip-environments and Poetry. But also understand why you want to use each one and why you prefer the one you do.

For simple tasks, I use pip with a virtual environment, I strive to be as consistent with creating environments as possible. I will work on using poetry more and more.

Data Science​

Python is the dominant working language for data science and machine learning.

  • Group by and Aggregation Guide Understanding how to filter and aggregate data can make your analysis so much faster. A lot of the things that are happening in meetings with non-technical people is that new hypotheses and assumptions are created, so being able to filter and aggregate quickly makes you able to almost on the spot do verification of said hypothesis. This can speed up the ideation phase of the data science life cycle considerably.
  • Surprise Libary for recomendation system

Web Development​

Boring setup for Python dependencies​

    1. Start project (mkdir, git init),
    1. Make virtualenv using virtualenvwrapper,
    1. Write project.toml file for setuptools,
    1. pip install -e .
    1. To add deps, add them to pyproject.toml and repeat step 4. Do not pip install deps directly. Do not pin deps to any particular version, but if you have to you can add constraints like >=5 (I need a feature introduced in v5).
    1. If you are writing a package to be pip installed by others then you're done. Read setuptools docs for how to build etc.
    1. If you also want to build an environment to run your code (e.g. docker image for deployment or serverless deployment etc) use pip-tools to pin your dependencies. (This is the only reason you need requirements.txt).
    1. For test dependencies (e.g. pytest) or dev dependencies (e.g. test server) leverage optional dependencies in the pyproject.toml file. This plays very nicely with tools like tox, which you should use. Use pre-commit for linting etc.

Links

Thoughts