refactor: do not require an empty struct in cross_validate to define …

src/api.rs

@@ -18,10 +18,6 @@ pub trait UnsupervisedEstimator<X, P> {
 
 /// An estimator for supervised learning, that provides method `fit` to learn from data and training values
 pub trait SupervisedEstimator<X, Y, P>: Predictor<X, Y> {
-    /// Empty constructor, instantiate an empty estimator. Object is dropped as soon as `fit()` is called.
-    /// used to pass around the correct `fit()` implementation.
-    /// by calling `::fit()`. mostly used to be used with `model_selection::cross_validate(...)`
-    fn new() -> Self;
     /// Fit a model to a training dataset, estimate model's parameters.
     /// * `x` - _NxM_ matrix with _N_ observations and _M_ features in each observation.
     /// * `y` - target training values of size _N_.
@@ -36,10 +32,6 @@ pub trait SupervisedEstimator<X, Y, P>: Predictor<X, Y> {
 /// In this one parameters are borrowed instead of moved, this is useful for parameters that carry
 /// references. Also to be used when there is no predictor attached to the estimator.
 pub trait SupervisedEstimatorBorrow<'a, X, Y, P> {
-    /// Empty constructor, instantiate an empty estimator. Object is dropped as soon as `fit()` is called.
-    /// used to pass around the correct `fit()` implementation.
-    /// by calling `::fit()`. mostly used to be used with `model_selection::cross_validate(...)`
-    fn new() -> Self;
     /// Fit a model to a training dataset, estimate model's parameters.
     /// * `x` - _NxM_ matrix with _N_ observations and _M_ features in each observation.
     /// * `y` - target training values of size _N_.

src/ensemble/random_forest_classifier.rs

@@ -196,14 +196,6 @@ impl<TX: Number + FloatNumber + PartialOrd, TY: Number + Ord, X: Array2<TX>, Y:
     SupervisedEstimator<X, Y, RandomForestClassifierParameters>
     for RandomForestClassifier<TX, TY, X, Y>
 {
-    fn new() -> Self {
-        Self {
-            parameters: Option::None,
-            trees: Option::None,
-            classes: Option::None,
-            samples: Option::None,
-        }
-    }
     fn fit(x: &X, y: &Y, parameters: RandomForestClassifierParameters) -> Result<Self, Failed> {
         RandomForestClassifier::fit(x, y, parameters)
     }

src/ensemble/random_forest_regressor.rs

@@ -175,14 +175,6 @@ impl<TX: Number + FloatNumber + PartialOrd, TY: Number, X: Array2<TX>, Y: Array1
     SupervisedEstimator<X, Y, RandomForestRegressorParameters>
     for RandomForestRegressor<TX, TY, X, Y>
 {
-    fn new() -> Self {
-        Self {
-            parameters: Option::None,
-            trees: Option::None,
-            samples: Option::None,
-        }
-    }
-
     fn fit(x: &X, y: &Y, parameters: RandomForestRegressorParameters) -> Result<Self, Failed> {
         RandomForestRegressor::fit(x, y, parameters)
     }

src/linear/elastic_net.rs

@@ -284,15 +284,6 @@ impl<TX: FloatNumber + RealNumber, TY: Number, X: Array2<TX>, Y: Array1<TY>> Par
 impl<TX: FloatNumber + RealNumber, TY: Number, X: Array2<TX>, Y: Array1<TY>>
     SupervisedEstimator<X, Y, ElasticNetParameters> for ElasticNet<TX, TY, X, Y>
 {
-    fn new() -> Self {
-        Self {
-            coefficients: Option::None,
-            intercept: Option::None,
-            _phantom_ty: PhantomData,
-            _phantom_y: PhantomData,
-        }
-    }
-
     fn fit(x: &X, y: &Y, parameters: ElasticNetParameters) -> Result<Self, Failed> {
         ElasticNet::fit(x, y, parameters)
     }

src/linear/lasso.rs

@@ -116,15 +116,6 @@ impl<TX: FloatNumber + RealNumber, TY: Number, X: Array2<TX>, Y: Array1<TY>> Par
 impl<TX: FloatNumber + RealNumber, TY: Number, X: Array2<TX>, Y: Array1<TY>>
     SupervisedEstimator<X, Y, LassoParameters> for Lasso<TX, TY, X, Y>
 {
-    fn new() -> Self {
-        Self {
-            coefficients: Option::None,
-            intercept: Option::None,
-            _phantom_ty: PhantomData,
-            _phantom_y: PhantomData,
-        }
-    }
-
     fn fit(x: &X, y: &Y, parameters: LassoParameters) -> Result<Self, Failed> {
         Lasso::fit(x, y, parameters)
     }

src/linear/linear_regression.rs

@@ -206,16 +206,6 @@ impl<
         Y: Array1<TY>,
     > SupervisedEstimator<X, Y, LinearRegressionParameters> for LinearRegression<TX, TY, X, Y>
 {
-    fn new() -> Self {
-        Self {
-            coefficients: Option::None,
-            intercept: Option::None,
-            solver: LinearRegressionParameters::default().solver,
-            _phantom_ty: PhantomData,
-            _phantom_y: PhantomData,
-        }
-    }
-
     fn fit(x: &X, y: &Y, parameters: LinearRegressionParameters) -> Result<Self, Failed> {
         LinearRegression::fit(x, y, parameters)
     }

src/linear/logistic_regression.rs

@@ -390,18 +390,6 @@ impl<TX: Number + FloatNumber + RealNumber, TY: Number + Ord, X: Array2<TX>, Y:
     SupervisedEstimator<X, Y, LogisticRegressionParameters<TX>>
     for LogisticRegression<TX, TY, X, Y>
 {
-    fn new() -> Self {
-        Self {
-            coefficients: Option::None,
-            intercept: Option::None,
-            classes: Option::None,
-            num_attributes: 0,
-            num_classes: 0,
-            _phantom_tx: PhantomData,
-            _phantom_y: PhantomData,
-        }
-    }
-
     fn fit(x: &X, y: &Y, parameters: LogisticRegressionParameters<TX>) -> Result<Self, Failed> {
         LogisticRegression::fit(x, y, parameters)
     }

src/linear/ridge_regression.rs

@@ -255,16 +255,6 @@ impl<
         Y: Array1<TY>,
     > SupervisedEstimator<X, Y, RidgeRegressionParameters<TX>> for RidgeRegression<TX, TY, X, Y>
 {
-    fn new() -> Self {
-        Self {
-            coefficients: Option::None,
-            intercept: Option::None,
-            solver: Option::None,
-            _phantom_ty: PhantomData,
-            _phantom_y: PhantomData,
-        }
-    }
-
     fn fit(x: &X, y: &Y, parameters: RidgeRegressionParameters<TX>) -> Result<Self, Failed> {
         RidgeRegression::fit(x, y, parameters)
     }

src/model_selection/mod.rs

@@ -91,8 +91,8 @@
 //!
 //! let cv = KFold::default().with_n_splits(3);
 //!
-//! let results = cross_validate(
-//!     LogisticRegression::new(),   //estimator
+//! // Need to pass generic type parameter to define the type of the estimator
+//! let results = cross_validate::<_,_,_,_,_,LogisticRegression<_,_,_,_>,_,_>(
 //!     &x, &y,                 //data
 //!     Default::default(),     //hyperparameters
 //!     &cv,                     //cross validation split
@@ -222,7 +222,6 @@ impl CrossValidationResult {
 /// * `cv` - the cross-validation splitting strategy, should be an instance of [`BaseKFold`](./trait.BaseKFold.html)
 /// * `score` - a metric to use for evaluation, see [metrics](../metrics/index.html)
 pub fn cross_validate<TX, TY, X, Y, H, E, K, S>(
-    _estimator: E, // just an empty placeholder to allow passing `fit()`
     x: &X,
     y: &Y,
     parameters: H,
@@ -271,7 +270,6 @@ where
 /// * `parameters` - parameters of selected estimator. Use `Default::default()` for default parameters.
 /// * `cv` - the cross-validation splitting strategy, should be an instance of [`BaseKFold`](./trait.BaseKFold.html)
 pub fn cross_val_predict<TX, TY, X, Y, H, E, K>(
-    _estimator: E, // just an empty placeholder to allow passing `fit()`
     x: &X,
     y: &Y,
     parameters: H,
@@ -360,9 +358,6 @@ mod tests {
         impl<X: Array2<f32>, Y: Array1<u32>, P: NoParameters> SupervisedEstimator<X, Y, P>
             for BiasedEstimator
         {
-            fn new() -> Self {
-                Self {}
-            }
             fn fit(_: &X, _: &Y, _: P) -> Result<BiasedEstimator, Failed> {
                 Ok(BiasedEstimator {})
             }
@@ -404,8 +399,7 @@ mod tests {
             ..KFold::default()
         };
 
-        let results = cross_validate(
-            BiasedEstimator {},
+        let results = cross_validate::<_, _, _, _, _, BiasedEstimator, _, _>(
             &x,
             &y,
             BiasedParameters {},
@@ -452,8 +446,7 @@ mod tests {
             ..KFold::default()
         };
 
-        let results = cross_validate(
-            KNNRegressor::new(),
+        let results = cross_validate::<_, _, _, _, _, KNNRegressor<_, _, _, _, _>, _, _>(
             &x,
             &y,
             Default::default(),
@@ -500,8 +493,7 @@ mod tests {
             ..KFold::default()
         };
 
-        let y_hat: Vec<f64> = cross_val_predict(
-            KNNRegressor::new(),
+        let y_hat: Vec<f64> = cross_val_predict::<_, _, _, _, _, KNNRegressor<_, _, _, _, _>, _>(
             &x,
             &y,
             KNNRegressorParameters::default()
@@ -544,8 +536,7 @@ mod tests {
 
         let cv = KFold::default().with_n_splits(3);
 
-        let results = cross_validate(
-            LogisticRegression::new(),
+        let results = cross_validate::<_, _, _, _, _, LogisticRegression<_, _, _, _>, _, _>(
             &x,
             &y,
             Default::default(),

src/naive_bayes/bernoulli.rs

@@ -381,13 +381,6 @@ impl<TX: Number + PartialOrd, TY: Number + Ord + Unsigned, X: Array2<TX>, Y: Arr
 impl<TX: Number + PartialOrd, TY: Number + Ord + Unsigned, X: Array2<TX>, Y: Array1<TY>>
     SupervisedEstimator<X, Y, BernoulliNBParameters<TX>> for BernoulliNB<TX, TY, X, Y>
 {
-    fn new() -> Self {
-        Self {
-            inner: Option::None,
-            binarize: Option::None,
-        }
-    }
-
     fn fit(x: &X, y: &Y, parameters: BernoulliNBParameters<TX>) -> Result<Self, Failed> {
         BernoulliNB::fit(x, y, parameters)
     }

src/naive_bayes/categorical.rs

@@ -347,12 +347,6 @@ pub struct CategoricalNB<T: Number + Unsigned, X: Array2<T>, Y: Array1<T>> {
 impl<T: Number + Unsigned, X: Array2<T>, Y: Array1<T>>
     SupervisedEstimator<X, Y, CategoricalNBParameters> for CategoricalNB<T, X, Y>
 {
-    fn new() -> Self {
-        Self {
-            inner: Option::None,
-        }
-    }
-
     fn fit(x: &X, y: &Y, parameters: CategoricalNBParameters) -> Result<Self, Failed> {
         CategoricalNB::fit(x, y, parameters)
     }

src/naive_bayes/gaussian.rs

@@ -291,12 +291,6 @@ impl<
         Y: Array1<TY>,
     > SupervisedEstimator<X, Y, GaussianNBParameters> for GaussianNB<TX, TY, X, Y>
 {
-    fn new() -> Self {
-        Self {
-            inner: Option::None,
-        }
-    }
-
     fn fit(x: &X, y: &Y, parameters: GaussianNBParameters) -> Result<Self, Failed> {
         GaussianNB::fit(x, y, parameters)
     }

src/naive_bayes/multinomial.rs

@@ -325,12 +325,6 @@ impl<TX: Number + Unsigned, TY: Number + Ord + Unsigned, X: Array2<TX>, Y: Array
 impl<TX: Number + Unsigned, TY: Number + Ord + Unsigned, X: Array2<TX>, Y: Array1<TY>>
     SupervisedEstimator<X, Y, MultinomialNBParameters> for MultinomialNB<TX, TY, X, Y>
 {
-    fn new() -> Self {
-        Self {
-            inner: Option::None,
-        }
-    }
-
     fn fit(x: &X, y: &Y, parameters: MultinomialNBParameters) -> Result<Self, Failed> {
         MultinomialNB::fit(x, y, parameters)
     }

src/neighbors/knn_classifier.rs

@@ -181,18 +181,6 @@ impl<TX: Number, TY: Number + Ord, X: Array2<TX>, Y: Array1<TY>, D: Distance<Vec
 impl<TX: Number, TY: Number + Ord, X: Array2<TX>, Y: Array1<TY>, D: Distance<Vec<TX>>>
     SupervisedEstimator<X, Y, KNNClassifierParameters<TX, D>> for KNNClassifier<TX, TY, X, Y, D>
 {
-    fn new() -> Self {
-        Self {
-            classes: Option::None,
-            y: Option::None,
-            knn_algorithm: Option::None,
-            weight: Option::None,
-            k: Option::None,
-            _phantom_tx: PhantomData,
-            _phantom_x: PhantomData,
-            _phantom_y: PhantomData,
-        }
-    }
     fn fit(x: &X, y: &Y, parameters: KNNClassifierParameters<TX, D>) -> Result<Self, Failed> {
         KNNClassifier::fit(x, y, parameters)
     }

src/neighbors/knn_regressor.rs

@@ -177,18 +177,6 @@ impl<TX: Number, TY: Number, X: Array2<TX>, Y: Array1<TY>, D: Distance<Vec<TX>>>
 impl<TX: Number, TY: Number, X: Array2<TX>, Y: Array1<TY>, D: Distance<Vec<TX>>>
     SupervisedEstimator<X, Y, KNNRegressorParameters<TX, D>> for KNNRegressor<TX, TY, X, Y, D>
 {
-    fn new() -> Self {
-        Self {
-            y: Option::None,
-            knn_algorithm: Option::None,
-            weight: Option::None,
-            k: Option::None,
-            _phantom_tx: PhantomData,
-            _phantom_ty: PhantomData,
-            _phantom_x: PhantomData,
-        }
-    }
-
     fn fit(x: &X, y: &Y, parameters: KNNRegressorParameters<TX, D>) -> Result<Self, Failed> {
         KNNRegressor::fit(x, y, parameters)
     }

src/svm/svc.rs

@@ -208,16 +208,6 @@ impl<TX: Number + RealNumber, TY: Number + Ord, X: Array2<TX>, Y: Array1<TY>> De
 impl<'a, TX: Number + RealNumber, TY: Number + Ord, X: Array2<TX>, Y: Array1<TY>>
     SupervisedEstimatorBorrow<'a, X, Y, SVCParameters<TX, TY, X, Y>> for SVC<'a, TX, TY, X, Y>
 {
-    fn new() -> Self {
-        Self {
-            classes: Option::None,
-            instances: Option::None,
-            parameters: Option::None,
-            w: Option::None,
-            b: Option::None,
-            phantomdata: PhantomData,
-        }
-    }
     fn fit(
         x: &'a X,
         y: &'a Y,

src/svm/svr.rs

@@ -177,15 +177,6 @@ impl<T: Number + FloatNumber + PartialOrd> Default for SVRParameters<T> {
 impl<'a, T: Number + FloatNumber + PartialOrd, X: Array2<T>, Y: Array1<T>>
     SupervisedEstimatorBorrow<'a, X, Y, SVRParameters<T>> for SVR<'a, T, X, Y>
 {
-    fn new() -> Self {
-        Self {
-            instances: Option::None,
-            parameters: Option::None,
-            w: Option::None,
-            b: T::zero(),
-            phantom: PhantomData,
-        }
-    }
     fn fit(x: &'a X, y: &'a Y, parameters: &'a SVRParameters<T>) -> Result<Self, Failed> {
         SVR::fit(x, y, parameters)
     }

src/tree/decision_tree_classifier.rs

@@ -507,19 +507,6 @@ impl<TX: Number + PartialOrd, TY: Number + Ord, X: Array2<TX>, Y: Array1<TY>>
     SupervisedEstimator<X, Y, DecisionTreeClassifierParameters>
     for DecisionTreeClassifier<TX, TY, X, Y>
 {
-    fn new() -> Self {
-        Self {
-            nodes: vec![],
-            parameters: Option::None,
-            num_classes: 0usize,
-            classes: vec![],
-            depth: 0u16,
-            _phantom_tx: PhantomData,
-            _phantom_x: PhantomData,
-            _phantom_y: PhantomData,
-        }
-    }
-
     fn fit(x: &X, y: &Y, parameters: DecisionTreeClassifierParameters) -> Result<Self, Failed> {
         DecisionTreeClassifier::fit(x, y, parameters)
     }

src/tree/decision_tree_regressor.rs

@@ -387,18 +387,6 @@ impl<TX: Number + PartialOrd, TY: Number, X: Array2<TX>, Y: Array1<TY>>
     SupervisedEstimator<X, Y, DecisionTreeRegressorParameters>
     for DecisionTreeRegressor<TX, TY, X, Y>
 {
-    fn new() -> Self {
-        Self {
-            nodes: vec![],
-            parameters: Option::None,
-            depth: 0u16,
-            _phantom_tx: PhantomData,
-            _phantom_ty: PhantomData,
-            _phantom_x: PhantomData,
-            _phantom_y: PhantomData,
-        }
-    }
-
     fn fit(x: &X, y: &Y, parameters: DecisionTreeRegressorParameters) -> Result<Self, Failed> {
         DecisionTreeRegressor::fit(x, y, parameters)
     }