Join GitHub today
GitHub is home to over 50 million developers working together to host and review code, manage projects, and build software together.
Sign upGitHub is where the world builds software
Millions of developers and companies build, ship, and maintain their software on GitHub — the largest and most advanced development platform in the world.
code: https://github.com/FlorinAndrei/c19 (clone this repo)
world.py and usa.py are exported directly from world.ipynb and usa.ipynb. The Jupyter notebooks are the "source of truth".
On macOS and Windows I can run the Jupyter notebooks without any issues.
On Linux (Ubuntu 16.04) I'm trying to run
python world.py. It's a headless system, but I've installed Xvfb. I've installed all dependencies via conda, except for fuzzyset which is installed via pip.Most of world.py works fine on Linux. Then it tries to spawn 4 workers, one for each CPU:
make_map() is in workers.py. It works on macOS / Win from the Jupyter notebook - it spawns multiple workers, multiple Orca processes. It fails on Linux (see error below). For Linux, I've tried to add
plotly.io.orca.ensure_server()- no luck.Do I need to start one Orca server per worker? On separate ports? Or what's causing the error?