feat: + cluster metrics

src/math/distance/euclidian.rs

@@ -29,6 +29,7 @@ use super::Distance;
 pub struct Euclidian {}
 
 impl Euclidian {
+    #[inline]
     pub(crate) fn squared_distance<T: RealNumber>(x: &Vec<T>, y: &Vec<T>) -> T {
         if x.len() != y.len() {
             panic!("Input vector sizes are different.");
@@ -36,7 +37,8 @@ impl Euclidian {
 
         let mut sum = T::zero();
         for i in 0..x.len() {
-            sum = sum + (x[i] - y[i]).powf(T::two());
+            let d = x[i] - y[i];
+            sum = sum + d * d;
         }
 
         sum

src/math/mod.rs

@@ -1,3 +1,4 @@
 /// Multitude of distance metrics are defined here
 pub mod distance;
 pub mod num;
+pub(crate) mod vector;

src/math/vector.rs

@@ -0,0 +1,40 @@
+use crate::math::num::RealNumber;
+use std::collections::HashMap;
+
+pub trait RealNumberVector<T: RealNumber> {
+    fn unique(&self) -> (Vec<T>, Vec<usize>);
+}
+
+impl<T: RealNumber> RealNumberVector<T> for Vec<T> {
+    fn unique(&self) -> (Vec<T>, Vec<usize>) {
+        let mut unique = self.clone();
+        unique.sort_by(|a, b| a.partial_cmp(b).unwrap());
+        unique.dedup();
+
+        let mut index = HashMap::with_capacity(unique.len());
+        for (i, u) in unique.iter().enumerate() {
+            index.insert(u.to_i64().unwrap(), i);
+        }
+
+        let mut unique_index = Vec::with_capacity(self.len());
+        for e in self {
+            unique_index.push(index[&e.to_i64().unwrap()]);
+        }
+
+        (unique, unique_index)
+    }
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+
+    #[test]
+    fn unique() {
+        let v1 = vec![0.0, 0.0, 1.0, 1.0, 2.0, 0.0, 4.0];
+        assert_eq!(
+            (vec!(0.0, 1.0, 2.0, 4.0), vec!(0, 0, 1, 1, 2, 0, 3)),
+            v1.unique()
+        );
+    }
+}