personal/time-to-botec
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
Simple estimation scripts which do the same in different programming languages.
32 Commits 1 Branch 0 Tags 27 MiB
C 57.3%
Makefile 24.1%
Nim 5.1%
OCaml 3.2%
Go 3.1%
Other 7%
8acdc283a2
Go to file
NunoSempere 8acdc283a2 look at the R code
2023-05-21 12:04:27 -04:00
C add fast output to C. 2023-05-21 12:02:53 -04:00
js tweak: make null window an object in order for this to run with bun 2022-12-07 19:14:31 +00:00
nim tweak: nim/hardcore -> nim/samples-from-scratch 2023-05-21 11:07:00 -04:00
python fix: improve warnings for a check which should never fail 2022-12-01 16:12:54 +00:00
R look at the R code 2023-05-21 12:04:27 -04:00
squiggle feat: rejiggle default number of samples. 2022-12-03 13:14:08 +00:00
wip/zig move nim to top level, add to README 2023-05-21 01:46:45 -04:00
.gitignore feat: add the node modules 2022-12-03 12:44:49 +00:00
README.md tweaks. 2023-05-21 01:54:03 -04:00
time.txt tweaks. 2023-05-21 01:54:03 -04:00

README.md

Time to BOTEC

About

This repository contains example of very simple code to manipulate samples in various programming languages. It implements this estimate:

p_a = 0.8
p_b = 0.5
p_c = p_a * p_b

dists = [0, 1, 1 to 3, 2 to 10] # each dist represented as 1M samples
weights = [(1 - p_c), p_c/2, p_c/4, p_c/4 ]

result = mixture(dists, weights)
mean(result)

As of now, it may be useful for checking the validity of simple estimations. The title of this repository is a pun on two meanings of "time to": "how much time does it take to do x", and "let's do x".

Current languages

  • C
  • Javascript (NodeJS)
  • Squiggle
  • R
  • Python
  • Nim

Performance table

With the time tool, using 1M samples:

Language Time
Nim 0m0.153s
C 0m0,442s
Node 0m0,732s
Squiggle 0m1,536s
R 0m7,000s
Python (CPython) 0m16,641s

I was very surprised that Node/Squiggle code was almost as fast as the raw C code. For the Python code, it's possible that the lack of speed is more a function of me not being as familiar with Python. It's also very possible that the code would run faster with PyPy.

I was also really happy with trying Nim. The version which beats all others is just the fastest "danger" compilation of Nim (the "release" compilation is 0m0.183s instead). The Nim version has the particularity that I define the normal function from scratch, using the Box–Muller transform. For Nim I also have a version of the code which takes around 4 seconds, where I define some very inefficient sine & logarithm functions to feed into the Box-Muller method, because it felt like fun to really write a botec tool really from scratch.

Languages I may add later

  • Julia (TuringML)
  • Rust
  • Lisp
  • Stan
  • Go
  • Zig
  • Forth
  • ... and suggestions welcome

Roadmap

The future of this project is uncertain. In most words, I simply forget about this repository.

To do:

  • Check whether the Squiggle code is producing 1M samples. Still not too sure.
  • Differentiate between initial startup time (e.g., compiling, loading environment) and runtime. This matters because startup time could be ~constant, so for larger projects only the runtime matters.

Other similar projects