Browse Source

Use batch_inversion_and_mul within native lagrange interpolation (#63)

master
Dev Ojha 3 years ago
committed by GitHub
parent
commit
02ee91d61b
No known key found for this signature in database GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 6 additions and 9 deletions
  1. +6
    -9
      src/poly/evaluations/univariate/lagrange_interpolator.rs

+ 6
- 9
src/poly/evaluations/univariate/lagrange_interpolator.rs

@ -1,5 +1,5 @@
use crate::poly::domain::vanishing_poly::VanishingPolynomial; use crate::poly::domain::vanishing_poly::VanishingPolynomial;
use ark_ff::{batch_inversion, PrimeField};
use ark_ff::{batch_inversion_and_mul, PrimeField};
use ark_std::vec::Vec; use ark_std::vec::Vec;
/// Struct describing Lagrange interpolation for a multiplicative coset I, /// Struct describing Lagrange interpolation for a multiplicative coset I,
/// with |I| a power of 2. /// with |I| a power of 2.
@ -71,21 +71,18 @@ impl LagrangeInterpolator {
- Z_{H}(t) = \prod_{j} (t-h*g^j) = (t^m-h^m), and - Z_{H}(t) = \prod_{j} (t-h*g^j) = (t^m-h^m), and
- v_{i} = 1 / \prod_{j \neq i} h(g^i-g^j). - v_{i} = 1 / \prod_{j \neq i} h(g^i-g^j).
Below we use the fact that v_{0} = 1/(m * h^(m-1)) and v_{i+1} = g * v_{i}. Below we use the fact that v_{0} = 1/(m * h^(m-1)) and v_{i+1} = g * v_{i}.
We compute the inverse of each coefficient, and then batch invert the entire result.
We first compute the inverse of each coefficient, except for the Z_H(t) term.
We then batch invert the entire result, and multiply by Z_H(t).
*/ */
let vp_t_inv = self
.domain_vp
.evaluate(&interpolation_point)
.inverse()
.unwrap();
let mut inverted_lagrange_coeffs: Vec<F> = Vec::with_capacity(self.all_domain_elems.len()); let mut inverted_lagrange_coeffs: Vec<F> = Vec::with_capacity(self.all_domain_elems.len());
for i in 0..self.domain_order { for i in 0..self.domain_order {
let l = vp_t_inv * self.v_inv_elems[i];
let l = self.v_inv_elems[i];
let r = self.all_domain_elems[i]; let r = self.all_domain_elems[i];
inverted_lagrange_coeffs.push(l * (interpolation_point - r)); inverted_lagrange_coeffs.push(l * (interpolation_point - r));
} }
let vp_t = self.domain_vp.evaluate(&interpolation_point);
let lagrange_coeffs = inverted_lagrange_coeffs.as_mut_slice(); let lagrange_coeffs = inverted_lagrange_coeffs.as_mut_slice();
batch_inversion::<F>(lagrange_coeffs);
batch_inversion_and_mul::<F>(lagrange_coeffs, &vp_t);
lagrange_coeffs.iter().cloned().collect() lagrange_coeffs.iter().cloned().collect()
} }

Loading…
Cancel
Save