Link Search Menu Expand Document

pytest + YourBase

On this page you’ll learn how to accelerate your pytest tests with YourBase. The quickest way to get started is to use a sample project; in this guide, we’ll use the open source project for the Unleash Python client.

Table of contents

  1. Introduction
  2. Prerequisites
  3. Enable acceleration for a sample project


YourBase Test Acceleration’s pytest hooks ensure that, by default:

  • It runs automatically on every invocation of pytest.
  • It runs without any additional configuration.
  • It runs irrespective of how you invoke the tests to run—be it via Makefile, Docker, or anything else—without any other setup.


  • Make sure tests are running successfully with pytest before installing YourBase.
  • Install YourBase.
  • Your project must use Git and must have at least one commit. Git must be installed.
  • Your Git workspace should be clean.

Enable acceleration for a sample project

Step 1: Set up the specified sample project on your machine as shown below:

Open a new shell prompt, and checkout this sample project from git

git clone

From your shell prompt, navigate to the directory where you checked out the project

cd unleash-client-python

Install dependencies

pip install -r requirements.txt

Step 2: Execute a command to run all the tests of the project

pytest tests

Let’s look at the output.

All messages that are prefixed with [YB] are logged by the YourBase Test Acceleration library. Note, that one of these messages says “No function-level dependency graph found; all tests will be run and traced.”—this is a cold run. During a cold run, YourBase Test Acceleration will build a dependency graph based on the tests that are executed. This dependency graph contains relationships between individual tests and the code they call.

Step 3: Without making any code change, execute the command to run all the tests again

pytest tests

Let’s look at the messages logged by YourBase Test Acceleration. One of the messages says: [YB] No code changed. Running only new tests, if any.”

Since you ran the tests without changing any code, YourBase Test Acceleration skipped all the tests to finish the run much more quickly than the last time—compare the time taken by both runs.

Step 4: Make a code change and execute the command to run all tests again as shown below:

Open tests/unit_tests/ in your text editor

vim tests/unit_tests/

Add the following print statement in the beginning of the method test_create_feature_true

print(Checking YourBase Test Acceleration after a code-change...)

Run the tests again using:

pytest tests

Let’s look at the output again.

If you modified the same function as above then your output will closely match the output above.

Look for the logs traced by YourBase Test Acceleration. You’ll see a message telling you how many functions have been altered and how many tests were affected. If you’re following the steps as is, you should see:

[YB] 1 function differs from the dependency graph
[YB] ~ tests/unit_tests/
[YB] Function-level dependency graph found 1 test affected

You can see form the logs that the existing dependency graph [link to How it works → Dependency graph section] was used to decide that only one test was affected by your code-change and that test was the only one that was run, while the remaining tests were skipped.

That’s it!

You just ran YourBase Test Acceleration on a project that uses the pytest testing framework. Now, you’re ready to use it in your own project.

Note that if your tests are going to take a while to run, you can run just a subset of your tests. Running a subset of tests will create a dependency graph just for those tests, so you can see YourBase Test Acceleration in action more quickly.