←back to thread

Show HN: Rust library for numerical integration of real-valued functions

(github.com)
123 points mtantaoui | 2 comments | 19 Nov 24 12:41 UTC | HN request time: 0.454s | source

Integrate is a fast, small, lightweight Rust library for performing numerical integration of real-valued functions. It is designed to integrate functions, providing a simple and efficient way to approximate definite integrals using various numerical methods.

Integrate supports a variety of numerical integration techniques: - Newton-Cotes methods:

  - Rectangle Rule.
  - Trapezoidal Rule.
  - Simpson's Rule.
  - Newton's 3/8 Rule.
- Gauss quadrature methods:

  - Gauss-Legendre.
  - Gauss-Laguerre.
  - Gauss-Hermite.
  - Gauss-Chebyshev First Kind.
  - Gauss-Chebyshev Second Kind.
- Adaptive Methods:

  - Adaptive Simpson's method
- Romberg’s method.
Show context
wjholden ◴[19 Nov 24 14:49 UTC] No.42184001[source]▶
I was always amazed that R can do:

  > integrate(dnorm, -Inf, +Inf)
  1 with absolute error < 9.4e-05
Can we do the same in this library?
replies(4): >>42184765 #>>42184870 #>>42185462 #>>42185618 #
1. Buttons840 ◴[19 Nov 24 15:48 UTC] No.42184765[source]▶
How many evaluations of the underlying function does it make? (Hoping someone will fire up their R interpreter and find out.)

Or, probably, dnorm is a probability distribution which includes a likeliness function, and a cumulative likeliness function, etc. I bet it doesn't work on arbitrary functions.

replies(1): >>42184883 #
2. thrasibule ◴[19 Nov 24 15:58 UTC] No.42184883[source]▶
R integrate is just a wrapper around quadpack. It works with arbitrary functions, but arguably dnorm is pretty well behaved.