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
 41
 42
 43
 44
 45
 46
 47
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105

#![warn(rust_2018_idioms)]
#![warn(missing_docs)]

//! # KNN
//!
//! `knn` provides for fast method of finding exact k nearest neighbours
//! for higher dimensional data. Also supports custom distance function.

mod heap;

#[derive(Clone)]
/// The core datastructure of the package.
/// Stores the points and the distance function.
pub struct PointCloud<'a, T> {
    // stores all points
    points: Vec<&'a T>,

    // return the distance given 2 ponints
    dist_fn: fn(&T, &T) -> f64,
}

impl<'a, T> PointCloud<'a, T> {
    /// Constructs a new `PointCloud<T>`.
    ///
    /// It accepts a function to compute distance between any 2 objects.
    ///
    /// # Examples
    /// ```
    /// extern crate knn;
    /// use knn::PointCloud;
    /// let manhattan_dist = |p: &[f64;2], q: &[f64;2]| -> f64 {(p[0] - q[0]).abs() + (p[1] - q[1]).abs()};
    /// let pc = PointCloud::new(manhattan_dist);
    /// ```
    pub fn new(dist_fn: fn(&T, &T) -> f64) -> Self {
        Self {
            points: Vec::new(),
            dist_fn: dist_fn,
        }
    }

    /// Adds a point to the PointCloud.
    ///
    /// It accepts a reference to an object.
    ///
    /// # Examples
    /// ```
    /// extern crate knn;
    /// use knn::PointCloud;
    /// let dummy_dist = |p: &[f64;2], q: &[f64;2]| -> f64 {0.0};
    /// let mut pc = PointCloud::new(dummy_dist);
    /// let p = [1.89, 5.63];
    /// pc.add_point(&p);
    /// ```
    pub fn add_point(&mut self, p: &'a T) {
        self.points.push(p)
    }

    /// Gets the k nearest neighbours of an object.
    ///
    /// It accepts a reference to the object and how many neighbours to return.
    ///
    /// # Example
    /// ```
    /// extern crate knn;
    /// use knn::PointCloud;
    /// let manhattan_dist = |p: &[f64;2], q: &[f64;2]| -> f64 {(p[0] - q[0]).abs() + (p[1] - q[1]).abs()};
    /// let mut pc = PointCloud::new(manhattan_dist);
    /// let points = vec![[1.0, 1.0], [2.0, 2.0], [3.0, 2.0]];
    /// for p in &points {
    ///     pc.add_point(&p);
    /// }
    /// let q = [0.5, 0.5];
    /// pc.get_nearest_k(&q, 2);
    ///
    /// ```
    pub fn get_nearest_k(&self, p: &T, k: usize) -> Vec<(f64, &T)> {
        let mut h = heap::Heap::new(k);
        for i in 0..self.points.len() {
            let d = (self.dist_fn)(self.points[i], p);
            h.insert(d, &self.points[i]);
        }
        let mut knn = Vec::new();
        for i in h.get_elements() {
            knn.push((i.0, *i.1));
        }
        knn
    }

    /// Get the len / number of objects in PointCloud
    /// Example:
    /// ```
    /// extern crate knn;
    /// use knn::PointCloud;
    /// let manhattan_dist = |p: &[f64;2], q: &[f64;2]| -> f64 {(p[0] - q[0]).abs() + (p[1] - q[1]).abs()};
    /// let mut pc = PointCloud::new(manhattan_dist);
    /// let points = vec![[1.0, 1.0], [2.0, 2.0], [3.0, 2.0]];
    /// for p in &points {
    ///     pc.add_point(&p);
    /// }
    /// pc.len();
    /// ```
    pub fn len(&self) -> usize {
        self.points.len()
    }
}