Error handling

Intermediate Python for Developers

George Boorman

Curriculum Manager, DataCamp

Pandas traceback

Traceback showing code that intentionally returns an error

except, raise
Try to anticipate how errors might occur

Design-thinking

How might people use our custom function?
Test these different approaches
Find what errors occur

Path created by walking across some grass to cut out a corner

¹ Image credit: https://www.flickr.com/photos/140641142@N05/

Error handling in custom functions

def average(values):
    # Calculate the average
    average_value = sum(values) / len(values)
    return average_value

Where might they go wrong?

Provide more than one argument
Use the wrong data type

Where might they go wrong?

sales_dict = {"cust_id": ["JL93", "MT12", "IY64"],
              "order_value": [43.21, 68.70, 82.19]}
average(sales_dict)

Traceback with TypeError

Error-handling techniques

Control flow if, elif, else
Docstrings

try-except

def average(values):

    try:

        # Code that might cause an error
        average_value = sum(values) / len(values)
        return average_value

    except:

        # Code to run if an error occurs
        print("average() accepts a list or set. Please provide a correct data type.")


average(sales_dict)

average() accepts a list or set. Please provide a correct data type.

raise

def average(values):
    # Check data type
    if type(values) in ["list", "set"]:

        # Run if appropriate data type was used
        average_value = sum(values) / len(values)
        return average_value

raise

def average(values):
    # Check data type
    if type(values) in ["list", "set"]:

        # Run if appropriate data type was used
        average_value = sum(values) / len(values)
        return average_value

    else:
        # Run if an Exception occurs
        raise

raise TypeError

def average(values):
    # Check data type
    if type(values) in ["list", "set"]:

        # Run if appropriate data type was used
        average_value = sum(values) / len(values)
        return average_value

    else:
        # Run if an Exception occurs
        raise TypeError("average() accepts a list or set, please provide a correct data type.")

raise TypeError output

average(sales_dict)

TypeError output showing the custom message to provide the correct data type

try-except vs. raise

`try`-`except`

Avoid errors being produced
Still execute subsequent code

`raise`

Will produce an error
Avoid executing subsequent code

Let's practice!

Intermediate Python for Developers