Merge branch 'development' into patch-0.3.2

.github/CONTRIBUTING.md

@@ -37,6 +37,8 @@ $ rust-code-analysis-cli -p src/algorithm/neighbour/fastpair.rs --ls 22 --le 213
 ```
 * find more information about what happens in your binary with [`twiggy`](https://rustwasm.github.io/twiggy/install.html). This need a compiled binary so create a brief `main {}` function using `smartcore` and then point `twiggy` to that file.
 
+* Please take a look to the output of a profiler to spot most evident performance problems, see [this guide about using a profiler](http://www.codeofview.com/fix-rs/2017/01/24/how-to-optimize-rust-programs-on-linux/).
+
 ## Issue Report Process
 
 1. Go to the project's issues.

Cargo.toml

@@ -2,7 +2,7 @@
 name = "smartcore"
 description = "Machine Learning in Rust."
 homepage = "https://smartcorelib.org"
-version = "0.4.0"
+version = "0.3.3"
 authors = ["smartcore Developers"]
 edition = "2021"
 license = "Apache-2.0"
@@ -48,7 +48,7 @@ getrandom = { version = "0.2.8", optional = true }
 wasm-bindgen-test = "0.3"
 
 [dev-dependencies]
-itertools = "0.10.5"
+itertools = "0.12.0"
 serde_json = "1.0"
 bincode = "1.3.1"
 

src/cluster/dbscan.rs

@@ -315,8 +315,7 @@ impl<TX: Number, TY: Number, X: Array2<TX>, Y: Array1<TY>, D: Distance<Vec<TX>>>
                         }
                     }
 
-                    while !neighbors.is_empty() {
-                        let neighbor = neighbors.pop().unwrap();
+                    while let Some(neighbor) = neighbors.pop() {
                         let index = neighbor.0;
 
                         if y[index] == outlier {

src/dataset/diabetes.rs

@@ -40,7 +40,7 @@ pub fn load_dataset() -> Dataset<f32, u32> {
         target: y,
         num_samples,
         num_features,
-        feature_names: vec![
+        feature_names: [
             "Age", "Sex", "BMI", "BP", "S1", "S2", "S3", "S4", "S5", "S6",
         ]
         .iter()

src/dataset/digits.rs

@@ -25,16 +25,14 @@ pub fn load_dataset() -> Dataset<f32, f32> {
         target: y,
         num_samples,
         num_features,
-        feature_names: vec![
-            "sepal length (cm)",
+        feature_names: ["sepal length (cm)",
             "sepal width (cm)",
             "petal length (cm)",
-            "petal width (cm)",
-        ]
+            "petal width (cm)"]
         .iter()
         .map(|s| s.to_string())
         .collect(),
-        target_names: vec!["setosa", "versicolor", "virginica"]
+        target_names: ["setosa", "versicolor", "virginica"]
             .iter()
             .map(|s| s.to_string())
             .collect(),

src/dataset/iris.rs

@@ -36,7 +36,7 @@ pub fn load_dataset() -> Dataset<f32, u32> {
         target: y,
         num_samples,
         num_features,
-        feature_names: vec![
+        feature_names: [
             "sepal length (cm)",
             "sepal width (cm)",
             "petal length (cm)",
@@ -45,7 +45,7 @@ pub fn load_dataset() -> Dataset<f32, u32> {
         .iter()
         .map(|s| s.to_string())
         .collect(),
-        target_names: vec!["setosa", "versicolor", "virginica"]
+        target_names: ["setosa", "versicolor", "virginica"]
             .iter()
             .map(|s| s.to_string())
             .collect(),

src/linalg/basic/arrays.rs

@@ -188,8 +188,7 @@ pub trait ArrayView1<T: Debug + Display + Copy + Sized>: Array<T, usize> {
                 _ => max,
             }
         };
-        self.iterator(0)
-            .fold(T::min_value(), |max, x| max_f(max, x))
+        self.iterator(0).fold(T::min_value(), max_f)
     }
     /// return min value from  the view
     fn min(&self) -> T
@@ -202,8 +201,7 @@ pub trait ArrayView1<T: Debug + Display + Copy + Sized>: Array<T, usize> {
                 _ => min,
             }
         };
-        self.iterator(0)
-            .fold(T::max_value(), |max, x| min_f(max, x))
+        self.iterator(0).fold(T::max_value(), min_f)
     }
     /// return the position of the max value of the view
     fn argmax(&self) -> usize

src/linalg/basic/matrix.rs

@@ -495,9 +495,9 @@ impl<T: Number + RealNumber> SVDDecomposable<T> for DenseMatrix<T> {}
 impl<'a, T: Debug + Display + Copy + Sized> Array<T, (usize, usize)> for DenseMatrixView<'a, T> {
     fn get(&self, pos: (usize, usize)) -> &T {
         if self.column_major {
-            &self.values[(pos.0 + pos.1 * self.stride)]
+            &self.values[pos.0 + pos.1 * self.stride]
         } else {
-            &self.values[(pos.0 * self.stride + pos.1)]
+            &self.values[pos.0 * self.stride + pos.1]
         }
     }
 
@@ -559,9 +559,9 @@ impl<'a, T: Debug + Display + Copy + Sized> ArrayView1<T> for DenseMatrixView<'a
 impl<'a, T: Debug + Display + Copy + Sized> Array<T, (usize, usize)> for DenseMatrixMutView<'a, T> {
     fn get(&self, pos: (usize, usize)) -> &T {
         if self.column_major {
-            &self.values[(pos.0 + pos.1 * self.stride)]
+            &self.values[pos.0 + pos.1 * self.stride]
         } else {
-            &self.values[(pos.0 * self.stride + pos.1)]
+            &self.values[pos.0 * self.stride + pos.1]
         }
     }
 
@@ -583,9 +583,9 @@ impl<'a, T: Debug + Display + Copy + Sized> MutArray<T, (usize, usize)>
 {
     fn set(&mut self, pos: (usize, usize), x: T) {
         if self.column_major {
-            self.values[(pos.0 + pos.1 * self.stride)] = x;
+            self.values[pos.0 + pos.1 * self.stride] = x;
         } else {
-            self.values[(pos.0 * self.stride + pos.1)] = x;
+            self.values[pos.0 * self.stride + pos.1] = x;
         }
     }
 
@@ -775,7 +775,7 @@ mod tests {
 
     #[test]
     fn test_from_iterator() {
-        let data = vec![1, 2, 3, 4, 5, 6];
+        let data = [1, 2, 3, 4, 5, 6];
 
         let m = DenseMatrix::from_iterator(data.iter(), 2, 3, 0);
 

src/linalg/basic/vector.rs

@@ -15,6 +15,25 @@ pub struct VecView<'a, T: Debug + Display + Copy + Sized> {
     ptr: &'a [T],
 }
 
+impl<T: Debug + Display + Copy + Sized> Array<T, usize> for &[T] {
+    fn get(&self, i: usize) -> &T {
+        &self[i]
+    }
+
+    fn shape(&self) -> usize {
+        self.len()
+    }
+
+    fn is_empty(&self) -> bool {
+        self.len() > 0
+    }
+
+    fn iterator<'b>(&'b self, axis: u8) -> Box<dyn Iterator<Item = &'b T> + 'b> {
+        assert!(axis == 0, "For one dimensional array `axis` should == 0");
+        Box::new(self.iter())
+    }
+}
+
 impl<T: Debug + Display + Copy + Sized> Array<T, usize> for Vec<T> {
     fn get(&self, i: usize) -> &T {
         &self[i]
@@ -47,6 +66,7 @@ impl<T: Debug + Display + Copy + Sized> MutArray<T, usize> for Vec<T> {
 }
 
 impl<T: Debug + Display + Copy + Sized> ArrayView1<T> for Vec<T> {}
+impl<T: Debug + Display + Copy + Sized> ArrayView1<T> for &[T] {}
 
 impl<T: Debug + Display + Copy + Sized> MutArrayView1<T> for Vec<T> {}
 
@@ -192,7 +212,7 @@ mod tests {
 
     #[test]
     fn test_len() {
-        let x = vec![1, 2, 3];
+        let x = [1, 2, 3];
         assert_eq!(3, x.len());
     }
 

src/linear/bg_solver.rs

@@ -162,7 +162,7 @@ mod tests {
         let a = DenseMatrix::from_2d_array(&[&[25., 15., -5.], &[15., 18., 0.], &[-5., 0., 11.]])
             .unwrap();
         let b = vec![40., 51., 28.];
-        let expected = vec![1.0, 2.0, 3.0];
+        let expected = [1.0, 2.0, 3.0];
 
         let mut x = Vec::zeros(3);
 

src/linear/logistic_regression.rs

@@ -898,11 +898,7 @@ mod tests {
 
         let y_hat = lr.predict(&x).unwrap();
 
-        let error: i32 = y
-            .into_iter()
-            .zip(y_hat.into_iter())
-            .map(|(a, b)| (a - b).abs())
-            .sum();
+        let error: i32 = y.into_iter().zip(y_hat).map(|(a, b)| (a - b).abs()).sum();
 
         assert!(error <= 1);
 

src/model_selection/hyper_tuning/grid_search.rs

@@ -3,9 +3,9 @@
 use crate::{
     api::{Predictor, SupervisedEstimator},
     error::{Failed, FailedError},
-    linalg::basic::arrays::{Array2, Array1},
-    numbers::realnum::RealNumber,
+    linalg::basic::arrays::{Array1, Array2},
     numbers::basenum::Number,
+    numbers::realnum::RealNumber,
 };
 
 use crate::model_selection::{cross_validate, BaseKFold, CrossValidationResult};

src/model_selection/kfold.rs

@@ -283,9 +283,7 @@ mod tests {
             (vec![0, 1, 2, 3, 7, 8, 9], vec![4, 5, 6]),
             (vec![0, 1, 2, 3, 4, 5, 6], vec![7, 8, 9]),
         ];
-        for ((train, test), (expected_train, expected_test)) in
-            k.split(&x).into_iter().zip(expected)
-        {
+        for ((train, test), (expected_train, expected_test)) in k.split(&x).zip(expected) {
             assert_eq!(test, expected_test);
             assert_eq!(train, expected_train);
         }
@@ -307,9 +305,7 @@ mod tests {
             (vec![0, 1, 2, 3, 7, 8, 9], vec![4, 5, 6]),
             (vec![0, 1, 2, 3, 4, 5, 6], vec![7, 8, 9]),
         ];
-        for ((train, test), (expected_train, expected_test)) in
-            k.split(&x).into_iter().zip(expected)
-        {
+        for ((train, test), (expected_train, expected_test)) in k.split(&x).zip(expected) {
             assert_eq!(test.len(), expected_test.len());
             assert_eq!(train.len(), expected_train.len());
         }

src/naive_bayes/mod.rs

@@ -40,7 +40,7 @@ use crate::linalg::basic::arrays::{Array1, Array2, ArrayView1};
 use crate::numbers::basenum::Number;
 #[cfg(feature = "serde")]
 use serde::{Deserialize, Serialize};
-use std::marker::PhantomData;
+use std::{cmp::Ordering, marker::PhantomData};
 
 /// Distribution used in the Naive Bayes classifier.
 pub(crate) trait NBDistribution<X: Number, Y: Number>: Clone {
@@ -92,11 +92,10 @@ impl<TX: Number, TY: Number, X: Array2<TX>, Y: Array1<TY>, D: NBDistribution<TX,
     /// Returns a vector of size N with class estimates.
     pub fn predict(&self, x: &X) -> Result<Y, Failed> {
         let y_classes = self.distribution.classes();
-        let (rows, _) = x.shape();
-        let predictions = (0..rows)
-            .map(|row_index| {
-                let row = x.get_row(row_index);
-                let (prediction, _probability) = y_classes
+        let predictions = x
+            .row_iter()
+            .map(|row| {
+                y_classes
                     .iter()
                     .enumerate()
                     .map(|(class_index, class)| {
@@ -106,11 +105,26 @@ impl<TX: Number, TY: Number, X: Array2<TX>, Y: Array1<TY>, D: NBDistribution<TX,
                                 + self.distribution.prior(class_index).ln(),
                         )
                     })
-                    .max_by(|(_, p1), (_, p2)| p1.partial_cmp(p2).unwrap())
-                    .unwrap();
-                *prediction
+                    // For some reason, the max_by method cannot use NaNs for finding the maximum value, it panics.
+                    // NaN must be considered as minimum values,
+                    // therefore it's like NaNs would not be considered for choosing the maximum value.
+                    // So we need to handle this case for avoiding panicking by using `Option::unwrap`.
+                    .max_by(|(_, p1), (_, p2)| match p1.partial_cmp(p2) {
+                        Some(ordering) => ordering,
+                        None => {
+                            if p1.is_nan() {
+                                Ordering::Less
+                            } else if p2.is_nan() {
+                                Ordering::Greater
+                            } else {
+                                Ordering::Equal
+                            }
+                        }
+                    })
+                    .map(|(prediction, _probability)| *prediction)
+                    .ok_or_else(|| Failed::predict("Failed to predict, there is no result"))
             })
-            .collect::<Vec<TY>>();
+            .collect::<Result<Vec<TY>, Failed>>()?;
         let y_hat = Y::from_vec_slice(&predictions);
         Ok(y_hat)
     }
@@ -119,3 +133,63 @@ pub mod bernoulli;
 pub mod categorical;
 pub mod gaussian;
 pub mod multinomial;
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+    use crate::linalg::basic::arrays::Array;
+    use crate::linalg::basic::matrix::DenseMatrix;
+    use num_traits::float::Float;
+
+    type Model<'d> = BaseNaiveBayes<i32, i32, DenseMatrix<i32>, Vec<i32>, TestDistribution<'d>>;
+
+    #[derive(Debug, PartialEq, Clone)]
+    struct TestDistribution<'d>(&'d Vec<i32>);
+
+    impl<'d> NBDistribution<i32, i32> for TestDistribution<'d> {
+        fn prior(&self, _class_index: usize) -> f64 {
+            1.
+        }
+
+        fn log_likelihood<'a>(
+            &'a self,
+            class_index: usize,
+            _j: &'a Box<dyn ArrayView1<i32> + 'a>,
+        ) -> f64 {
+            match self.0.get(class_index) {
+                &v @ 2 | &v @ 10 | &v @ 20 => v as f64,
+                _ => f64::nan(),
+            }
+        }
+
+        fn classes(&self) -> &Vec<i32> {
+            &self.0
+        }
+    }
+
+    #[test]
+    fn test_predict() {
+        let matrix = DenseMatrix::from_2d_array(&[&[1, 2, 3], &[4, 5, 6], &[7, 8, 9]]);
+
+        let val = vec![];
+        match Model::fit(TestDistribution(&val)).unwrap().predict(&matrix) {
+            Ok(_) => panic!("Should return error in case of empty classes"),
+            Err(err) => assert_eq!(
+                err.to_string(),
+                "Predict failed: Failed to predict, there is no result"
+            ),
+        }
+
+        let val = vec![1, 2, 3];
+        match Model::fit(TestDistribution(&val)).unwrap().predict(&matrix) {
+            Ok(r) => assert_eq!(r, vec![2, 2, 2]),
+            Err(_) => panic!("Should success in normal case with NaNs"),
+        }
+
+        let val = vec![20, 2, 10];
+        match Model::fit(TestDistribution(&val)).unwrap().predict(&matrix) {
+            Ok(r) => assert_eq!(r, vec![20, 20, 20]),
+            Err(_) => panic!("Should success in normal case without NaNs"),
+        }
+    }
+}

src/neighbors/knn_regressor.rs

@@ -298,7 +298,7 @@ mod tests {
             DenseMatrix::from_2d_array(&[&[1., 2.], &[3., 4.], &[5., 6.], &[7., 8.], &[9., 10.]])
                 .unwrap();
         let y: Vec<f64> = vec![1., 2., 3., 4., 5.];
-        let y_exp = vec![1., 2., 3., 4., 5.];
+        let y_exp = [1., 2., 3., 4., 5.];
         let knn = KNNRegressor::fit(
             &x,
             &y,
@@ -326,7 +326,7 @@ mod tests {
             DenseMatrix::from_2d_array(&[&[1., 2.], &[3., 4.], &[5., 6.], &[7., 8.], &[9., 10.]])
                 .unwrap();
         let y: Vec<f64> = vec![1., 2., 3., 4., 5.];
-        let y_exp = vec![2., 2., 3., 4., 4.];
+        let y_exp = [2., 2., 3., 4., 4.];
         let knn = KNNRegressor::fit(&x, &y, Default::default()).unwrap();
         let y_hat = knn.predict(&x).unwrap();
         assert_eq!(5, Vec::len(&y_hat));

src/preprocessing/categorical.rs

@@ -281,7 +281,7 @@ mod tests {
     )]
     #[test]
     fn hash_encode_f64_series() {
-        let series = vec![3.0, 1.0, 2.0, 1.0];
+        let series = [3.0, 1.0, 2.0, 1.0];
         let hashable_series: Vec<CategoricalFloat> =
             series.iter().map(|v| v.to_category()).collect();
         let enc = CategoryMapper::from_positional_category_vec(hashable_series);