spacetracks
spacetracks on Hacker News
It defaults to 25x. At 1x it looks like nothing is happening.
Former LeoLabs engineer ama
I'm interested in applying transforms to vector tile coordinates at render time. What libraries do you recommend for that? I have zero experience with map rendering but I understand coordinate transformation well. I'd like to render a log azimuthal earth. I know I'll need to pull tiles at different scales and stitch them together, just not sure which library to use to make that easy.
Posted by
spacetracks on Wednesday July 19, 2023 @11:14AM
from the show-hn dept.
https://pypi.org/project/chonkify/
I created chonker to avoid indexing errors from writing index%chunsize. I considered using the standard approach of returning data chunks but find it more difficult to read for my common use case of checkpointing. Chonker returns a chonk data structure that can be used to make decisions when processing a chunk of data.
Posted by
spacetracks on Sunday July 17, 2022 @12:25PM
from the pypi.org dept.
"The second factor, and probably the more damning one, is that most ML codes don't actually use that much dynamism." I would argue that this in true precisely because it is not available in an AD system. When I tell friends and coworkers about what zygote can do they light up and start describing different use cases they have that could benefit from AD. Diff eq solving is a big one.
Qubits are susceptible to all kinds of environmental noise. That's why current designs are cooled to cryogenic temperatures. Certain kinds of Majorana particles solve this problem by making the qubit distributed. In a handwaving way, the 1 state of the qubit is on one end of a wire, and the zero state is on the other; really there is a Majorana particle on each end of the wire, and by manipulating them you can control the state of the qubit. To manipulate the qubit you must interact with both ends of the wire at the same time. That goes for noise as well. And noise typically acts locally, not in two places at the same time, so the qubit is protected.
Came here to say something like this. One big problem in the field is detection. They are detected only indirectly using current methods. And there are other particles that could trigger a detection. Majorana particles have interesting properties in theory, but these properties have yet to be validated. The most interesting, and useful for quantum computing, is the ability for a pair of Majorana Fermions to "record", in the form of a qubit, how many times they have been wrapped around each other. But this requires having two, and being able to move them at will. Every new material we find them in is a chance for success.