Demonstration of powerfull yaml syntax of taskfiles for job automation, environment building and dependency resolution
More on https://taskfile.dev.
Please install task from
https://github.com/go-task/task/releases/tag/v2.8.0
-
Windows.
- Unarchive it
- put to C:\Windows\System32 or other Dir in %PATH% environment variable
-
Linux amd Mac OSs.
- unarchive it:
tar xvf task*.tar.gz - make runnable:
chmod +x task - move it to $PATH dir:
mv task /usr/local/bin
- unarchive it:
For local development we have need to manage python virtual environments, and sometimes python versions. There are a couple ways of doing it:
python -m venv+ manual python verion installpyenv(manual install) +python -m venvpyenv(manual install) +Pipenv- poetry + pyenv https://github.com/python-poetry/poetry
- conda https://docs.conda.io/en/latest/license.html
Well most painfull part here is not the venv management,
which can be reduced to python 3 module venv python -m venv,
but the python itself version.
Currently in 2020 we still have ubuntu 16 with default python 3.5,
different kinds of mac os with brew python as the only alternative,
and manuall install or use of choko on windows.
The most "python way" solution is currently poetry,
(which still requires python version management).
The most simple and easier for data scientiets
approach is conda. which have a lots of precompiled modules
on mac and win and lin.
My Personal choice is wide python management by conda, and package management with poetry.
- Install miniconda
- Install poetry
curl -sSL https://raw.githubusercontent.com/python-poetry/poetry/master/get-poetry.py | python(Invoke-WebRequest -Uri https://raw.githubusercontent.com/python-poetry/poetry/master/get-poetry.py -UseBasicParsing).Content | python
For more environment independent parts its very handy to use to use docker
On windows download from https://hub.docker.com/editions/community/docker-ce-desktop-windows/