Error Handling

What is an Exception?

In programming, an exception is an event that disrupts the normal sequential flow of execution. When Python encounters an error at runtime — such as dividing by zero, reading a missing file, or applying an operation to the wrong type — it raises an exception object. This object propagates up the call stack until it reaches code that explicitly handles it. If nothing handles it, the program terminates and prints a traceback.

In Data Science, data is messy and unpredictable. A division by zero, a missing file, or wrong data types will crash your entire pipeline if you don't handle exceptions properly.

The Try/Except Block

try:
    # Risky code
    result = 10 / 0
except ZeroDivisionError:
    # What to do if it fails
    print("Cannot divide by zero!")
finally:
    # Always runs (cleanup)
    print("Execution finished.")

Common Data Science Exceptions

• KeyError: Accessing a missing dictionary key or Pandas column.
• TypeError: Applying math to strings.
• IndexError: Accessing a list index that doesn't exist.
• ValueError: Converting a non-numeric string to float.

💡 Best Practice: Never use a "bare except" (just except:). Always catch the specific error so you don't accidentally hide bugs.