cargo fmt

Author: maxprogrammer007

algorithms/linfa-trees/examples/iris_random_forest.rs

@@ -10,13 +10,11 @@ fn main() -> Result<(), Box<dyn std::error::Error>> {
     let mut rng = thread_rng();
 
     // 1. Load, shuffle, and split the Iris dataset (80% train, 20% valid)
-    let (train, valid) = iris()
-        .shuffle(&mut rng)
-        .split_with_ratio(0.8);
+    let (train, valid) = iris().shuffle(&mut rng).split_with_ratio(0.8);
 
     // 2. Single‐tree baseline
     let dt_model = DecisionTree::params()
-        .max_depth(None)    // no depth limit
+        .max_depth(None) // no depth limit
         .fit(&train)?;
     let dt_preds = dt_model.predict(valid.records.clone());
     let dt_cm = dt_preds.confusion_matrix(&valid)?;
@@ -26,7 +24,7 @@ fn main() -> Result<(), Box<dyn std::error::Error>> {
     let rf_model = RandomForestParams::new(50)
         .max_depth(Some(5))
         .feature_subsample(0.7)
-        .seed(42)           // fix RNG seed for reproducibility
+        .seed(42) // fix RNG seed for reproducibility
         .fit(&train)?;
     let rf_preds = rf_model.predict(valid.records.clone());
     let rf_cm = rf_preds.confusion_matrix(&valid)?;

algorithms/linfa-trees/src/decision_trees/mod.rs

@@ -7,4 +7,4 @@ pub use algorithm::*;
 pub use hyperparams::*;
 pub use iter::*;
 pub use tikz::*;
-pub mod random_forest;
+pub mod random_forest;

algorithms/linfa-trees/src/decision_trees/random_forest.rs

@@ -98,15 +98,10 @@ fn bootstrap<F: Float>(
     Dataset::new(rec, tgt)
 }
 
-impl<F: Float + Send + Sync> Fit<Array2<F>, Array1<usize>, Error>
-    for RandomForestValidParams<F>
-{
+impl<F: Float + Send + Sync> Fit<Array2<F>, Array1<usize>, Error> for RandomForestValidParams<F> {
     type Object = RandomForestClassifier<F>;
 
-    fn fit(
-        &self,
-        dataset: &DatasetBase<Array2<F>, Array1<usize>>,
-    ) -> Result<Self::Object, Error> {
+    fn fit(&self, dataset: &DatasetBase<Array2<F>, Array1<usize>>) -> Result<Self::Object, Error> {
         let mut rng = StdRng::seed_from_u64(self.seed);
         let mut trees = Vec::with_capacity(self.n_trees);
         let mut feats_list = Vec::with_capacity(self.n_trees);
@@ -141,9 +136,7 @@ impl<F: Float + Send + Sync> Fit<Array2<F>, Array1<usize>, Error>
     }
 }
 
-impl<F: Float> Predict<Array2<F>, Array1<usize>>
-    for RandomForestClassifier<F>
-{
+impl<F: Float> Predict<Array2<F>, Array1<usize>> for RandomForestClassifier<F> {
     fn predict(&self, x: Array2<F>) -> Array1<usize> {
         let n = x.nrows();
         // adjust 100 to the expected number of classes if known
@@ -162,7 +155,8 @@ impl<F: Float> Predict<Array2<F>, Array1<usize>>
         Array1::from(
             (0..n)
                 .map(|i| {
-                    votes.iter()
+                    votes
+                        .iter()
                         .enumerate()
                         .max_by_key(|(_, v)| v[i])
                         .map(|(lbl, _)| lbl)

algorithms/linfa-trees/tests/random_forest.rs

@@ -10,9 +10,7 @@ use rand::SeedableRng;
 fn iris_random_forest_high_accuracy() {
     // reproducible split
     let mut rng = StdRng::seed_from_u64(42);
-    let (train, valid) = iris()
-        .shuffle(&mut rng)
-        .split_with_ratio(0.8);
+    let (train, valid) = iris().shuffle(&mut rng).split_with_ratio(0.8);
 
     let model = RandomForestParams::new(100)
         .max_depth(Some(10))