feat: Make SerDe optional

src/linear/elastic_net.rs

@@ -56,7 +56,7 @@
 //! <script id="MathJax-script" async src="https://cdn.jsdelivr.net/npm/mathjax@3/es5/tex-mml-chtml.js"></script>
 use std::fmt::Debug;
 
-use serde::{Deserialize, Serialize};
+#[cfg(feature = "serde")] use serde::{Deserialize, Serialize};
 
 use crate::api::{Predictor, SupervisedEstimator};
 use crate::error::Failed;
@@ -67,7 +67,8 @@ use crate::math::num::RealNumber;
 use crate::linear::lasso_optimizer::InteriorPointOptimizer;
 
 /// Elastic net parameters
-#[derive(Serialize, Deserialize, Debug, Clone)]
+#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
+#[derive(Debug, Clone)]
 pub struct ElasticNetParameters<T: RealNumber> {
     /// Regularization parameter.
     pub alpha: T,
@@ -84,7 +85,8 @@ pub struct ElasticNetParameters<T: RealNumber> {
 }
 
 /// Elastic net
-#[derive(Serialize, Deserialize, Debug)]
+#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
+#[derive(Debug)]
 pub struct ElasticNet<T: RealNumber, M: Matrix<T>> {
     coefficients: M,
     intercept: T,

src/linear/lasso.rs

@@ -24,7 +24,7 @@
 //! <script id="MathJax-script" async src="https://cdn.jsdelivr.net/npm/mathjax@3/es5/tex-mml-chtml.js"></script>
 use std::fmt::Debug;
 
-use serde::{Deserialize, Serialize};
+#[cfg(feature = "serde")] use serde::{Deserialize, Serialize};
 
 use crate::api::{Predictor, SupervisedEstimator};
 use crate::error::Failed;
@@ -34,7 +34,8 @@ use crate::linear::lasso_optimizer::InteriorPointOptimizer;
 use crate::math::num::RealNumber;
 
 /// Lasso regression parameters
-#[derive(Serialize, Deserialize, Debug, Clone)]
+#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
+#[derive(Debug, Clone)]
 pub struct LassoParameters<T: RealNumber> {
     /// Controls the strength of the penalty to the loss function.
     pub alpha: T,
@@ -47,7 +48,8 @@ pub struct LassoParameters<T: RealNumber> {
     pub max_iter: usize,
 }
 
-#[derive(Serialize, Deserialize, Debug)]
+#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
+#[derive(Debug)]
 /// Lasso regressor
 pub struct Lasso<T: RealNumber, M: Matrix<T>> {
     coefficients: M,

src/linear/linear_regression.rs

@@ -62,14 +62,15 @@
 //! <script id="MathJax-script" async src="https://cdn.jsdelivr.net/npm/mathjax@3/es5/tex-mml-chtml.js"></script>
 use std::fmt::Debug;
 
-use serde::{Deserialize, Serialize};
+#[cfg(feature = "serde")] use serde::{Deserialize, Serialize};
 
 use crate::api::{Predictor, SupervisedEstimator};
 use crate::error::Failed;
 use crate::linalg::Matrix;
 use crate::math::num::RealNumber;
 
-#[derive(Serialize, Deserialize, Debug, Clone)]
+#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
+#[derive(Debug, Clone)]
 /// Approach to use for estimation of regression coefficients. QR is more efficient but SVD is more stable.
 pub enum LinearRegressionSolverName {
     /// QR decomposition, see [QR](../../linalg/qr/index.html)
@@ -79,14 +80,16 @@ pub enum LinearRegressionSolverName {
 }
 
 /// Linear Regression parameters
-#[derive(Serialize, Deserialize, Debug, Clone)]
+#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
+#[derive(Debug, Clone)]
 pub struct LinearRegressionParameters {
     /// Solver to use for estimation of regression coefficients.
     pub solver: LinearRegressionSolverName,
 }
 
 /// Linear Regression
-#[derive(Serialize, Deserialize, Debug)]
+#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
+#[derive(Debug)]
 pub struct LinearRegression<T: RealNumber, M: Matrix<T>> {
     coefficients: M,
     intercept: T,

src/linear/logistic_regression.rs

@@ -56,7 +56,7 @@ use std::cmp::Ordering;
 use std::fmt::Debug;
 use std::marker::PhantomData;
 
-use serde::{Deserialize, Serialize};
+#[cfg(feature = "serde")] use serde::{Deserialize, Serialize};
 
 use crate::api::{Predictor, SupervisedEstimator};
 use crate::error::Failed;
@@ -68,11 +68,13 @@ use crate::optimization::line_search::Backtracking;
 use crate::optimization::FunctionOrder;
 
 /// Logistic Regression parameters
-#[derive(Serialize, Deserialize, Debug, Clone)]
+#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
+#[derive(Debug, Clone)]
 pub struct LogisticRegressionParameters {}
 
 /// Logistic Regression
-#[derive(Serialize, Deserialize, Debug)]
+#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
+#[derive(Debug)]
 pub struct LogisticRegression<T: RealNumber, M: Matrix<T>> {
     coefficients: M,
     intercept: M,

src/linear/ridge_regression.rs

@@ -58,7 +58,7 @@
 //! <script id="MathJax-script" async src="https://cdn.jsdelivr.net/npm/mathjax@3/es5/tex-mml-chtml.js"></script>
 use std::fmt::Debug;
 
-use serde::{Deserialize, Serialize};
+#[cfg(feature = "serde")] use serde::{Deserialize, Serialize};
 
 use crate::api::{Predictor, SupervisedEstimator};
 use crate::error::Failed;
@@ -66,7 +66,8 @@ use crate::linalg::BaseVector;
 use crate::linalg::Matrix;
 use crate::math::num::RealNumber;
 
-#[derive(Serialize, Deserialize, Debug, Clone)]
+#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
+#[derive(Debug, Clone)]
 /// Approach to use for estimation of regression coefficients. Cholesky is more efficient but SVD is more stable.
 pub enum RidgeRegressionSolverName {
     /// Cholesky decomposition, see [Cholesky](../../linalg/cholesky/index.html)
@@ -76,7 +77,8 @@ pub enum RidgeRegressionSolverName {
 }
 
 /// Ridge Regression parameters
-#[derive(Serialize, Deserialize, Debug, Clone)]
+#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
+#[derive(Debug, Clone)]
 pub struct RidgeRegressionParameters<T: RealNumber> {
     /// Solver to use for estimation of regression coefficients.
     pub solver: RidgeRegressionSolverName,
@@ -88,7 +90,8 @@ pub struct RidgeRegressionParameters<T: RealNumber> {
 }
 
 /// Ridge regression
-#[derive(Serialize, Deserialize, Debug)]
+#[cfg_attr(feature = "serde", derive(Serialize, Deserialize))]
+#[derive(Debug)]
 pub struct RidgeRegression<T: RealNumber, M: Matrix<T>> {
     coefficients: M,
     intercept: T,