cadic/nmesh repository overview

⁠

NMesh

Modules⁠ • Code Design⁠ • Code Structure⁠ • How To Use⁠ • Docker⁠ • PythonEnv⁠ • Ressources⁠ •

NMesh is a Python package that provides two high-level features:

• A simple Mesh processor
• A list of tool to convert mesh files into point cloud

You can reuse your favorite Python packages such as NumPy, SciPy and Cython to extend ZakuroCache integration.

⁠Modules

At a granular level, NMesh is a library that consists of the following components:

⁠Code design

• We recommend using Docker for dev and production. Therefore we encourage its usage all other the repo.
• We have vanilla and sandbox environment.
  • Vanilla refers to a prebuilt docker image that already contains system dependencies.
  • Sandbox referes a predbuilt docker image that contains the code of this repo.
• Semantic versioning https://semver.org/⁠ . We commit fix to a.b.x, features to a.x.c and stable release (master) to x.b.c.
• PR are done to dev and reviewed for additional features. This should only be reviewed by the engineers in the team.
• PR are done to master for official (internal) release of the codes. This should be reviewed by the maximum number of engineers.
• The ETL jobs are scatter accross sequential refinement of the data landing/bronze/silver/gold
• Modules and scripts: Any piece of code that can of use for parent module modules should be moved at a higher level.
  • eg: functional.py contains common funtions for etl.bronze and etl.silver

...
├── etl
│   ├── bronze
│   │   ├── __init__.py
│   │   └── __main__.py
│   ├── functional.py
│   ├── __init__.py
│   └── landing
│       ├── __init__.py
│       └── __main__.py
├── functional.py
├── __init__.py
...

• Modules should ideally contain a __main__.py that demo an exeution of the module
  • etl/bronze/__main__.py describes an etl job for the creation of the bronze parition
  • trainer/__main__.py describes the training pipeline

⁠Code structure

from setuptools import setup
from nmesh import __version__
setup(
    name='nmesh',
    version=__version__,
    packages=[
        "nmesh",
        "nmesh.core",
        "nmesh.cp"
    ],
    url='https://github.com/JeanMaximilienCadic/nmesh',
    include_package_data=True,
    package_data={"": ["*.yml"]},
    long_description="".join(open("README.md", "r").readlines()),
    long_description_content_type='text/markdown',
    license='MIT',
    author='Jean Maximilien Cadic',
    python_requires='>=3.6',
    install_requires=[r.rsplit()[0] for r in open("requirements.txt")],
    author_email='[email protected]',
    description='GNU Tools for python',
    classifiers=[
        "Programming Language :: Python :: 3.6",
        "License :: OSI Approved :: MIT License",
    ]
)

⁠How to use

To clone and run this application, you'll need Git⁠ and https://docs.docker.com/docker-for-mac/install/⁠ and Python installed on your computer. From your command line:

Install the package:

# Clone this repository and install the code
git clone https://github.com/JeanMaximilienCadic/nmesh

# Go into the repository
cd nmesh

⁠Makefile

Exhaustive list of make commands:

install_wheels
sandbox_cpu
sandbox_gpu
build_sandbox
push_environment
push_container_sandbox
push_container_vanilla
pull_container_vanilla
pull_container_sandbox
build_vanilla
clean
build_wheels
auto_branch 

⁠Docker

(* recommended)

To build and run the docker image

make build
make docker_run_sandbox_cpu

⁠PythonEnv

(* not recommended)

make install_wheels

⁠Ressources

• Vanilla: https://en.wikipedia.org/wiki/Vanilla_software⁠
• Sandbox: https://en.wikipedia.org/wiki/Sandbox_(software_development)⁠
• All you need is docker: https://www.theregister.com/2014/05/23/google_containerization_two_billion/⁠
• Dev in containers : https://code.visualstudio.com/docs/remote/containers⁠
• Delta lake partitions: https://k21academy.com/microsoft-azure/data-engineer/delta-lake/⁠