update linear_regression.py

Author: PurvalBhude

machine_learning/linear_regression.py

@@ -23,21 +23,15 @@ def collect_dataset():
             "master/Week1/ADRvsRating.csv",
             timeout=10,
         )
-        response.raise_for_status()  # Check for HTTP errors
+        response.raise_for_status()  # Raise an error for failed HTTP requests
         lines = response.text.splitlines()
-        data = []
-        for item in lines:
-            item = item.split(",")
-            data.append(item)
-        data.pop(0)  # This is for removing the labels from the list
+        data = [line.split(",") for line in lines]
+        data.pop(0)  # Remove the labels from the list
         dataset = np.matrix(data)
         return dataset
     except requests.exceptions.RequestException as e:
-        print(f"Error in fetching dataset: {e}")
-        return None
-    except Exception as e:
-        print(f"Unexpected error in processing dataset: {e}")
-        return None
+        print(f"Error fetching dataset: {e}")
+        return None  # Return None if dataset fetching fails
 
 
 def run_steep_gradient_descent(data_x, data_y, len_data, alpha, theta):
@@ -51,15 +45,13 @@ def run_steep_gradient_descent(data_x, data_y, len_data, alpha, theta):
                        curr_features - alpha_ * gradient(w.r.t. feature)
     """
     try:
-        n = len_data
-        prod = np.dot(theta, data_x.transpose())
-        prod -= data_y.transpose()
+        prod = np.dot(theta, data_x.transpose()) - data_y.transpose()
         sum_grad = np.dot(prod, data_x)
-        theta = theta - (alpha / n) * sum_grad
+        theta = theta - (alpha / len_data) * sum_grad
         return theta
     except Exception as e:
-        print(f"Error during gradient descent: {e}")
-        return None
+        print(f"Error in gradient descent: {e}")
+        return theta  # Return current theta even if an error occurs
 
 
 def sum_of_square_error(data_x, data_y, len_data, theta):
@@ -71,14 +63,13 @@ def sum_of_square_error(data_x, data_y, len_data, theta):
     :return          : sum of square error computed from given feature's
     """
     try:
-        prod = np.dot(theta, data_x.transpose())
-        prod -= data_y.transpose()
+        prod = np.dot(theta, data_x.transpose()) - data_y.transpose()
         sum_elem = np.sum(np.square(prod))
         error = sum_elem / (2 * len_data)
         return error
     except Exception as e:
         print(f"Error in calculating sum of square error: {e}")
-        return None
+        return float('inf')  # Return infinity in case of an error
 
 
 def run_linear_regression(data_x, data_y):
@@ -87,31 +78,20 @@ def run_linear_regression(data_x, data_y):
     :param data_y  : contains the output (result vector)
     :return        : feature for line of best fit (Feature vector)
     """
-    try:
-        iterations = 100000
-        alpha = 0.0001550
-
-        no_features = data_x.shape[1]
-        len_data = data_x.shape[0]
-
-        theta = np.zeros((1, no_features))
+    iterations = 100000
+    alpha = 0.0001550
+    len_data = data_x.shape[0] - 1
+    no_features = data_x.shape[1]
+    theta = np.zeros((1, no_features))
 
+    try:
         for i in range(iterations):
             theta = run_steep_gradient_descent(data_x, data_y, len_data, alpha, theta)
-            if theta is None:  # If gradient descent fails, exit
-                print("Gradient descent failed. Exiting.")
-                return None
             error = sum_of_square_error(data_x, data_y, len_data, theta)
-            if error is None:  # If error calculation fails, exit
-                print("Error calculation failed. Exiting.")
-                return None
-            if i % 1000 == 0:  # Print every 1000 iterations
-                print(f"At Iteration {i + 1} - Error is {error:.5f}")
-
-        return theta
+            print(f"At Iteration {i + 1} - Error is {error:.5f}")
     except Exception as e:
-        print(f"Error in linear regression: {e}")
-        return None
+        print(f"Error during linear regression: {e}")
+    return theta
 
 
 def mean_absolute_error(predicted_y, original_y):
@@ -125,14 +105,14 @@ def mean_absolute_error(predicted_y, original_y):
         return total / len(original_y)
     except Exception as e:
         print(f"Error in calculating mean absolute error: {e}")
-        return None
+        return float('inf')
 
 
 def main():
-    """Driver function"""
+    """Driver function."""
     data = collect_dataset()
     if data is None:
-        print("Failed to collect or process the dataset. Exiting.")
+        print("Failed to retrieve dataset. Exiting.")
         return
 
     try:
@@ -141,16 +121,12 @@ def main():
         data_y = data[:, -1].astype(float)
 
         theta = run_linear_regression(data_x, data_y)
-        if theta is None:
-            print("Linear regression failed. Exiting.")
-            return
-
         len_result = theta.shape[1]
-        print("Resultant Feature vector:")
+        print("Resultant Feature vector : ")
         for i in range(len_result):
             print(f"{theta[0, i]:.5f}")
     except Exception as e:
-        print(f"Unexpected error in main: {e}")
+        print(f"Error in main execution: {e}")
 
 
 if __name__ == "__main__":