Gaussian Naive Bayes (GaussianNB)

Cargo.toml

@@ -71,6 +71,7 @@ members = [
     "linfa-svm",
     "linfa-hierarchical",
     "linfa-ica",
+    "linfa-bayes",
     "datasets",
 ]
 

README.md

@@ -36,6 +36,7 @@ Where does `linfa` stand right now? [Are we learning yet?](http://www.arewelearn
 | [trees](linfa-trees/) | Decision trees | Experimental  | Supervised learning | Linear decision trees
 | [svm](linfa-svm/) | Support Vector Machines | Tested  | Supervised learning | Classification or regression analysis of labeled datasets | 
 | [hierarchical](linfa-hierarchical/) | Agglomerative hierarchical clustering | Tested | Unsupervised learning | Cluster and build hierarchy of clusters |
+| [bayes](linfa-bayes/) | Naive Bayes | Tested | Supervised learning | Contain's Gaussian Naive Bayes |
 
 We believe that only a significant community effort can nurture, build, and sustain a machine learning ecosystem in Rust - there is no other way forward.
 

linfa-bayes/Cargo.toml

@@ -0,0 +1,21 @@
+[package]
+name = "linfa-bayes"
+version = "0.1.0"
+authors = ["VasanthakumarV <vasanth260m12@gmail.com>"]
+description = "Collection of Naive Bayes Algorithms"
+edition = "2018"
+license = "MIT/Apache-2.0"
+repository = "https://github.com/rust-ml/linfa"
+readme = "README.md"
+keywords = ["factorization", "machine-learning", "linfa", "unsupervised"]
+categories = ["algorithms", "mathematics", "science"]
+
+[dependencies]
+ndarray = { version = "0.13" , features = ["blas", "approx"]}
+ndarray-stats = "0.3"
+linfa = { version = "0.2.1", path = ".." }
+
+[dev-dependencies]
+approx = "0.3"
+linfa = { path = ".." }
+linfa-datasets = { version = "0.2.0", path = "../datasets", features = ["winequality"] }

linfa-bayes/README.md

@@ -0,0 +1,27 @@
+# Naive Bayes
+
+`linfa-bayes` aims to provide pure Rust implementations of Naive Bayes algorithms. 
+
+## The Big Picture
+
+`linfa-bayes` is a crate in the [`linfa`](https://crates.io/crates/linfa) ecosystem, an effort to create a toolkit for classical Machine Learning implemented in pure Rust, akin to Python's `scikit-learn`.
+
+## Current state
+
+`linfa-bayes` currently provides an implementation of the following methods: 
+
+- Gaussian Naive Bayes (GaussianNB)
+
+## Examples
+
+There is an usage example in the `examples/` directory. To run, use:
+
+```bash
+$ cargo run --example winequality 
+```
+
+## License
+Dual-licensed to be compatible with the Rust project.
+
+Licensed under the Apache License, Version 2.0 <http://www.apache.org/licenses/LICENSE-2.0> or the MIT license <http://opensource.org/licenses/MIT>, at your option. This file may not be copied, modified, or distributed except according to those terms.
+

linfa-bayes/examples/winequality.rs

@@ -0,0 +1,39 @@
+use std::error::Error;
+
+use linfa::metrics::ToConfusionMatrix;
+use linfa::traits::{Fit, Predict};
+use linfa_bayes::GaussianNbParams;
+
+fn main() -> Result<(), Box<dyn Error>> {
+    // Wine with rating greater than 6 is considered good
+    fn tag_classes(x: &usize) -> usize {
+        if *x > 6 {
+            1
+        } else {
+            0
+        }
+    };
+
+    // Read in the dataset and convert continuous target into categorical
+    let data = linfa_datasets::winequality().map_targets(tag_classes);
+
+    let (train, valid) = data.split_with_ratio_view(0.9);
+
+    // Train the model
+    let model = GaussianNbParams::params().fit(&train)?;
+
+    let pred = model.predict(valid.records);
+
+    // Construct confusion matrix
+    let cm = pred.confusion_matrix(&valid);
+
+    // classes    | 1          | 0
+    // 1          | 10         | 12
+    // 0          | 7          | 130
+    //
+    // accuracy 0.8805031, MCC 0.45080978
+    println!("{:?}", cm);
+    println!("accuracy {}, MCC {}", cm.accuracy(), cm.mcc());
+
+    Ok(())
+}

linfa-bayes/src/error.rs

@@ -0,0 +1,27 @@
+use std::fmt;
+
+use ndarray_stats::errors::MinMaxError;
+
+pub type Result<T> = std::result::Result<T, BayesError>;
+
+#[derive(Debug)]
+pub enum BayesError {
+    /// Error when performing Max operation on data
+    Stats(MinMaxError),
+}
+
+impl fmt::Display for BayesError {
+    fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
+        match self {
+            Self::Stats(error) => write!(f, "Ndarray Stats Error: {}", error),
+        }
+    }
+}
+
+impl From<MinMaxError> for BayesError {
+    fn from(error: MinMaxError) -> Self {
+        Self::Stats(error)
+    }
+}
+
+impl std::error::Error for BayesError {}