Add integration test for notebooks #389
Conversation
renamed all mentions of urls to uris to enable s3 support
…-print-statements Feature/50/logger instead of print statements
…data-types-for-major-use-cases added common data types
…r-tasks fixed broken cluster tests
…ask-if-fails handled error in case learning task, deleting original task
…erving added basic ray serve examples
feat(datalayer): format
Fix minor bugs and typos minor refactors logger to logging formatting and refactors
Fix some bugs
docs(cdc): Added working example on cdc added integration tests refactors format
Fix some issues and tests Fix minor typo Refactor tests and cdc code Fix minor typos and doc string
|
Great PR summary, @thejumpman2323! FYI, in case it helps (I think this is still WiP?) the same type of workflow in
It's interesting how you do the same thing, but from within Question: would it be cleaner if it is its own CI element (like in @thejumpman2323 call logic: GHA -> pytest -> python -> jupyter -> python |
Hi, |
| completed_process = subprocess.run(['python3', os.path.join(tmp_dir, py_file_path)], capture_output=True, text=True, timeout=10) | ||
| assert completed_process.returncode == 0 | ||
|
|
||
| def test_notebooks(): |
There was a problem hiding this comment.
Would prefer this to be configured manually - some notebooks won't work without credentials, etc..
Could we do just the mnist initially?
There was a problem hiding this comment.
We would need a test account for openAI, and secrets management on GitHub.
rec
force-pushed
the
main
branch
2 times, most recently
from
bf5f0c5 to
0f516a0
Compare
Solves a task in #135
This pull request introduces an integration test that automates the conversion of Jupyter notebooks to Python files and runs them individually as tests. This integration test enhances our testing infrastructure by providing a streamlined process for verifying the functionality and correctness of Jupyter notebook conversions.
Changes Made
Added a new integration test module test_notebooks.py which performs the Jupyter notebook to Python file conversion and execution.
Implemented the necessary logic to convert Jupyter notebooks to Python files using the nbconvert library.
Configured the test module to execute the converted Python files as individual tests using the testing framework.
Test Execution Flow
The integration test module test_notebooks.py locates the target Jupyter notebook files.
Each notebook file is automatically converted to a corresponding Python file using the nbconvert library.
The converted Python files are executed individually as tests, ensuring their correctness and functionality.
Test results, including any failures or errors, are reported in the test execution summary.
Benefits
Automation: The integration test automates the process of converting Jupyter notebooks to Python files and executing them as tests.
Verification: This test ensures that Jupyter notebook conversions produce valid Python code that can be executed successfully.
Continuous Integration: The integration test can be incorporated into our CI/CD pipeline to catch potential issues early in the development process.
Improved Code Reliability: Running Jupyter notebooks as tests helps identify any code inconsistencies or errors during conversion, leading to more reliable code.
This integration test contributes to the overall quality and reliability of our codebase, enabling us to confidently convert Jupyter notebooks to Python files and validate their correctness through automated testing.