machine_learning: accept numpy.matrix inputs and add test

Author: Mahadev-05

machine_learning/ridge_regression.py

@@ -26,7 +26,7 @@
 from __future__ import annotations
 
 from dataclasses import dataclass
-from typing import Optional
+
 
 import httpx
 import numpy as np
@@ -62,10 +62,11 @@ def __post_init__(self) -> None:
 
     @staticmethod
     def _add_intercept(features: np.ndarray) -> np.ndarray:
-        if features.ndim != 2:
+        arr = np.asarray(features, dtype=float)
+        if arr.ndim != 2:
             raise ValueError("features must be a 2D array")
-        n_samples = features.shape[0]
-        return np.c_[np.ones(n_samples), features]
+        n_samples = arr.shape[0]
+        return np.c_[np.ones(n_samples), arr]
 
     def fit(
         self, features: np.ndarray, target: np.ndarray, add_intercept: bool = True
@@ -81,6 +82,9 @@ def fit(
         add_intercept: bool
             If True the model will add a bias column of ones to `features`.
         """
+        features = np.asarray(features, dtype=float)
+        target = np.asarray(target, dtype=float)
+
         if features.ndim != 2:
             raise ValueError("features must be a 2D array")
         if target.ndim != 1:
@@ -119,6 +123,8 @@ def predict(self, features: np.ndarray, add_intercept: bool = True) -> np.ndarra
         """
         if self.weights is None:
             raise ValueError("Model is not trained")
+
+        features = np.asarray(features, dtype=float)
         x = features if not add_intercept else self._add_intercept(features)
         return x @ self.weights
 

machine_learning/tests/test_ridge_regression.py

@@ -49,3 +49,17 @@ def test_input_validation():
     model = RidgeRegression()
     with pytest.raises(ValueError):
         model.fit(np.array([1, 2, 3]), np.array([1, 2, 3]))
+
+
+def test_accepts_numpy_matrix():
+    from machine_learning.ridge_regression import collect_dataset
+
+    data = collect_dataset()
+    X = np.c_[data[:, 0].astype(float)]  # numpy.matrix
+    y = np.ravel(data[:, 1].astype(float))
+
+    model = RidgeRegression(learning_rate=0.0002, lambda_=0.01, epochs=500)
+    model.fit(X, y)
+    preds = model.predict(X)
+    assert preds.shape == (y.shape[0],)
+    assert mean_absolute_error(preds, y) >= 0.0