[Packaging] Please publish a PyPI-hosted stable release. about pyeit HOT 6 CLOSED

@leycec it can now be installed via

$ pip install pyeit

the conda-forge recipes PR has been submitted too.
conda-forge/staged-recipes#18025

from pyeit.

Thank you for your support :)

I've been using conda these days and try to publish on conda-forge first, you can help me by commiting some PRs!

pyEIT is a starting code for us EIT users. Currently I do not think it is production ready as a python package (It is yet lacking a clear and clean API interfaces for both EIT algorithm development and EIT application GUIs). I am working on this improvement these days and learned a lot from other fem packages.

from pyeit.

Sorry for the late delay. I am now manage (actually start to learn how) to publish pyeit on pypi and conda-forge.

from pyeit.

Oh, sure! That's great to hear. EIT is such an exciting scientific endeavour – and pyEIT is at the vanguard of that endeavour in the Python space.

Please do take your time to stabilize a well-designed public API. I'm sure everyone will leap on-board once pyEIT kicks out its first stable release. And thanks again. 😄

from pyeit.

Uh-oh. I've got some good news and I've got bad news, @liubenyuan. The good news is that I'm currently packaging both pyEIT and py-dbar for Gentoo Linux, because both projects are awesome and deserve greater visibility.

The bad news is that py-dbar requires pyEIT as a mandatory runtime dependency. Moreover, I was both astonished and frightened to find that py-dbar actively imports from pyEIT's unstable API for various meshing and FEM-related functionality: e.g.,

# In "py_dbar/sim.py":
import pyeit.mesh as mesh
from pyeit.mesh import quality
from pyeit.eit.utils import eit_scan_lines
from pyeit.eit.fem import Forward

😨 😱

from pyeit.

Big congratulations! Incredibly proud of your hard work, @liubenyuan. Thanks so much for investing into the conda-forge ecosystem, too. This makes pyEIT so much easier to integrate for everyone involved.

We can't wait to see where you take pyEIT next. 😄

from pyeit.