imgal 0.3.1

A fast and open-source scientific image processing and algorithm library.
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40

use imgal::distribution::{inverse_normal_cdf, normalized_gaussian};
use imgal::integration::midpoint;
use imgal::prelude::*;

const TOLERANCE: f64 = 1e-10;
const THREADS: Option<usize> = Some(0);

fn approx_equal(a: f64, b: f64) -> bool {
    (a - b).abs() < TOLERANCE
}

/// Tests that `inverse_normal_cdf` returns the expected values for known
/// probabilites and boundary cases.
#[test]
fn distribution_inverse_normal_cdf_expected_results() -> Result<(), ImgalError> {
    assert_eq!(inverse_normal_cdf(0.1)?, -1.2815515641401563);
    assert_eq!(inverse_normal_cdf(0.975)?, 1.959963986120195);
    assert_eq!(inverse_normal_cdf(0.5)?, 0.0);
    assert_eq!(inverse_normal_cdf(0.0)?, f64::NEG_INFINITY);
    assert_eq!(inverse_normal_cdf(1.0)?, f64::INFINITY);
    Ok(())
}

/// Tests that `normalized_gaussian` returns the expected results for index
/// `100` and the distribution integrates to approximately `1.0`.
#[test]
fn distribution_normalized_gaussian_expected_results() {
    let gauss_arr_a_par = normalized_gaussian(0.5, 256, 0.4, 2.0, THREADS);
    let gauss_arr_b_par = normalized_gaussian(2.0, 256, 4.0, 2.0, THREADS);
    let gauss_arr_a_seq = normalized_gaussian(0.5, 256, 0.4, 2.0, None);
    let gauss_arr_b_seq = normalized_gaussian(2.0, 256, 4.0, 2.0, None);
    assert!(approx_equal(midpoint(&gauss_arr_a_par, None, None), 1.0));
    assert!(approx_equal(midpoint(&gauss_arr_b_par, None, None), 1.0));
    assert!(approx_equal(midpoint(&gauss_arr_a_seq, None, None), 1.0));
    assert!(approx_equal(midpoint(&gauss_arr_b_seq, None, None), 1.0));
    assert!(approx_equal(gauss_arr_a_par[100], 0.0021260086));
    assert!(approx_equal(gauss_arr_b_par[100], 0.0044655072));
    assert!(approx_equal(gauss_arr_a_seq[100], 0.0021260086));
    assert!(approx_equal(gauss_arr_b_seq[100], 0.0044655072));
}