here's the code for anyone curious: http://git.jb55.com/umap-playground/file/harps_umap.py.html
I'm always amazed at how succinct python datascience libs are...
oh neat umap can learn an embedding using neural networks: https://umap-learn.readthedocs.io/en/latest/parametric_umap.html
I should try this
I wanted to see if UMAP could cluster radial velocity in the HARPS data without including the radial velocity in the training set. I'm not sure what the other parameters even mean and I had no motivation for doing this other than to see what would happen. look kinda cool at least
"The beauty [of the Dirac equation] boils down to the properties of spinor representations of the Lorentz group. [..] In other words, you should comprehensively learn and analyze the theory of representations of Lie groups and Lie algebras."
fzf has become some kind of universal interface for me. my `f` alias is a fzf command launcher which launches other fzf commands. also made it similar to bash's ctrl-r so there is a history of executed commands.
its a chonky boy
per-process bandwidth monitoring with bpftrace https://www.gcardone.net/2020-07-31-per-process-bandwidth-monitoring-on-Linux-with-bpftrace/
my node monitors my cold-store addresses in watch-only mode. so if someone breaks in there is nothing to steal. not only can you be your own bank, but you can be more secure than them as well. especially with multisig setups.
earth actually might have been flat(ish) at an early point in its development. It may explain why the earth and moon are of such different composition
dev, nix, linux, rust, c, gamedev, protocols, metaverse, physics, dad
Bitcoin Maston Instance