Unit testing with

pytest
INTRODUCTION TO TESTING IN PYTHON

Alexander Levin
Data Scientist
What is a unit test
Unit - the smallest working part that can be tested.

Examples of units: functions, methods, classes, modules, etc.

Unit testing - software testing method.

Unit testing allows one to scrutinize the correctness of a unit.

Test case - a set of unit inputs and expected outputs.

Test case summarizes a particular piece of the problem.

INTRODUCTION TO TESTING IN PYTHON

Why to use unit tests
Unit tests - are a foundation for testing "the bigger picture" of the software.

Use cases:

When bugs found

During development

After implemented changes

INTRODUCTION TO TESTING IN PYTHON

How to create a unit test
Step-by-step:

1. Decide which units to test

2. Define test cases (the creative part):

"What are the possible unit outcomes?"

"How can one use the unit?"

"How should the unit behave in all those cases?"

3. Write code for each test case

4. Run the tests and analyze the results

INTRODUCTION TO TESTING IN PYTHON

Creating a unit test: example
Unit to test:

# Function for a sum of elements

def sum_of_arr(array:list) -> int:
return sum(array)

Test cases:

1. Input is a list of numbers (as expected) - should return the sum

2. Input is an empty list - should return 0

3. Input is a list containing one number - should return the number

INTRODUCTION TO TESTING IN PYTHON

Creating a unit test: code
# Test Case 1: regular array
def test_regular():
assert sum_of_arr([1, 2, 3]) == 6
assert sum_of_arr([100, 150]) == 250

# Test Case 2: empty list

def test_empty():
assert sum_of_arr([]) == 0

# Test Case 3: one number

def test_one_number():
assert sum_of_arr([10]) == 10
assert sum_of_arr([0]) == 0

INTRODUCTION TO TESTING IN PYTHON

Summary
Unit test - a test that verifies that a unit works as expected.

Test case - inputs and outputs summarizing a particular piece of the problem.

Use cases:

When bugs found

During development

After implemented changes

Defining test cases - involves creativity.

INTRODUCTION TO TESTING IN PYTHON

 Let's practice!
INTRODUCTION TO TESTING IN PYTHON
 Feature testing with
pytest
INTRODUCTION TO TESTING IN PYTHON

Alexander Levin
Data Scientist
What is a feature test
Feature Examples:
Data distribution check
a software system functionality.
Report preparation
satisfies a particular user's requirement.
Features

are wider that units.

End-user can use features.

Feature testing

software testing method.

verify the behavior of a specific feature.

INTRODUCTION TO TESTING IN PYTHON

Units vs. features: personal computer
Units: Features:

Each individual button Keyboard

Pixels on the screen Screen

LED diodes Illumination

Blocks on the disk File system

INTRODUCTION TO TESTING IN PYTHON

Why do we use feature tests
Feature testing helps:

To test things at the scope of the user interaction with a system.

The scope is wider than units:

One unit does not work - does not mean the feature is NOT OK.

Vice versa:

All units work as expected - does not mean the feature is OK.

INTRODUCTION TO TESTING IN PYTHON

Feature test example: setup
Setup and defining the feature code:

# Setup
import pandas as pd
import pytest

df = pd.read_csv('laptops.csv')

# Filter feature
def filter_data_by_manuf(df, manufacturer_name):
filtered_df = df\
[df["Manufacturer"] == manufacturer_name]
return filtered_df

INTRODUCTION TO TESTING IN PYTHON

Feature test example: testing
Testing code:

# Feature test function

def test_unique():
manuf_name = 'Apple'
filtered = filter_data_by_manuf(df, manuf_name)
assert filtered\
['Manufacturer'].nunique() == 1
assert filtered\
['Manufacturer'].unique() == [manuf_name]

INTRODUCTION TO TESTING IN PYTHON

Summary
Feature testing - software testing method, allows to verify the behavior of a specific feature.

Features are wider that units:

If a button is a unit, then the keyboard is a feature.

Feature testing helps to ensure that users get exactly what they expect.

To create feature tests one has to create test cases.

The key to design feature tests - to understand what the features are in a specific system.

INTRODUCTION TO TESTING IN PYTHON

 Let's practice!
INTRODUCTION TO TESTING IN PYTHON
 Integration testing
with pytest
INTRODUCTION TO TESTING IN PYTHON

Alexander Levin
Data Scientist
What is integration testing?
Integration testing - software testing method, that allows to verify that an interaction
behaves normally.

Integration - an interaction between 2 or more modules inside of a system.

INTRODUCTION TO TESTING IN PYTHON

Integrations in real-life projects
Examples:

Power cable

Internet connection

File reading driver

Database connection

Application Programming Interface (API) integration

INTRODUCTION TO TESTING IN PYTHON

What can go wrong
Potential integration problems:

Lost connection

Loss of data
Interaction delays

Low bandwidth

Version conflicts

Interface mismatch

Others

INTRODUCTION TO TESTING IN PYTHON

Example of integration testing
import pytest, os

@pytest.fixture
def setup_file():
# Create temporary file
file = "test_file.txt"
with open(file, "w") as f1:
f1.write("Test data 1")
yield file
os.remove(file)

def test_fs(setup_file):
file = setup_file
# Check that the file was created successfully
assert os.path.exists(file)

INTRODUCTION TO TESTING IN PYTHON

Summary
Integration testing - software testing method; allows to verify that an integration works as
expected.

Real-life projects include a lot of different integrations in them.

Integration testing helps to prevent lots of potential problems.

Example: checking integration between Python and a file system.

INTRODUCTION TO TESTING IN PYTHON

 Let's practice!
INTRODUCTION TO TESTING IN PYTHON
 Performance testing
with pytest
INTRODUCTION TO TESTING IN PYTHON

Alexander Levin
Data Scientist
What is performance testing
Performance - how efficiently does software utilizes the resources of the system to
accomplish a task.

Performance Testing - is a type of testing that measures software performance.

INTRODUCTION TO TESTING IN PYTHON

When performance testing is important
Resources: Cases:

Execution Time Website speed optimization

CPU App receiving millions of requests

RAM Path planning for a robot vacuum

Other resources

INTRODUCTION TO TESTING IN PYTHON

Benchmark fixture
Installation:

pip install pytest-benchmark

# Example_1.py
import time
def test_func(benchmark):
benchmark(time.sleep, 1)

CLI Command:

pytest Example_1.py

INTRODUCTION TO TESTING IN PYTHON

Benchmarking results
The results we get after we execute the CLI Command:

For time.sleep(3) instead of time.sleep(1) :

INTRODUCTION TO TESTING IN PYTHON

Benchmark decorator
# Example_2.py
import time
def test_func(benchmark):
@benchmark
def sleep_for_1_sec():
time.sleep(1)

CLI Command:

pytest Example_2.py

INTRODUCTION TO TESTING IN PYTHON

Summary
Performance testing - a type of testing that measures the software performance.
Resources are usually finite.

Helpful when resources are constrained.

We can use pytest-benchmark fixture by:

calling benchmark directly

using @benchmark as a decorator

The results describe the sample of measured runs in seconds.

INTRODUCTION TO TESTING IN PYTHON

 Let's practice!
INTRODUCTION TO TESTING IN PYTHON