Skip to content

Latest commit

 

History

History
122 lines (96 loc) · 5.43 KB

.testing.md

File metadata and controls

122 lines (96 loc) · 5.43 KB

Testing

We are using some tools for testing modifications and new features additions.

  • automated tests using travis and pytest.
  • coverage.py, a tool for measuring code coverage of Python programs.

Links

PyTest

Next we'll explain the steps to test your code. When a test is run, a . means that the test passed and an F means it failed.

  1. In your environment, install pytest by running pip install pytest or conda install -c conda-forge pytest

  2. Now, simply run pytest, and it will give you and output like this if all tests passed

++ pytest --showlocals --pyargs pymove/
============================================== test session starts ==============================================
platform linux -- Python 3.7.6, pytest-5.4.2, py-1.8.1, pluggy-0.13.1
rootdir: /home/flycher/Documentos/PyMove, inifile: setup.cfg
collected 95 items

pymove/tests/test_init.py .                                                                                [  1%]
pymove/tests/core/test_data_frame.py ......................................................................[ 74%]
pymove/tests/preprocessing/test_compression.py .                                                           [ 75%]
pymove/tests/preprocessing/test_filters.py ...............                                                 [ 91%]
pymove/tests/preprocessing/test_segmentation.py .....                                                      [ 96%]
pymove/tests/preprocessing/test_stay_point_detection.py ...                                                [100%]

====================================== 95 passed, 1 warning in 3.62s =============================================
  1. And like this if any tests fail
++ pytest --showlocals --pyargs pymove/
============================================== test session starts ==============================================
platform linux -- Python 3.7.6, pytest-5.4.2, py-1.8.1, pluggy-0.13.1
rootdir: /home/flycher/Documentos/PyMove, inifile: setup.cfg
collected 95 items

pymove/tests/test_init.py .                                                                                [  1%]
pymove/tests/core/test_data_frame.py ......................................................................[ 74%]
pymove/tests/preprocessing/test_compression.py .                                                           [ 75%]
pymove/tests/preprocessing/test_filters.py .F.............                                                 [ 91%]
pymove/tests/preprocessing/test_segmentation.py .....                                                      [ 96%]
pymove/tests/preprocessing/test_stay_point_detection.py ...                                                [100%]

=================================================================================================================
 def test_by_datetime():
        move_df, cols = _prepare_df_default(list_data_1)

        filter_values_start = filters.by_datetime(
            move_df, start_datetime='2008-10-23 05:53:11')
        expected = DataFrame(data=[
            [39.984199952392578, 116.31932067871094, Timestamp('2008-10-23 05:53:06'), 1],
            [39.984222412109375, 116.31940460205078, Timestamp('2008-10-23 05:53:11'), 2],
            [39.984222412109375, 116.31940460205078, Timestamp('2008-10-23 05:53:11'), 2]
        ], columns=cols, index=[1, 2, 3]
        )
>       assert_frame_equal(filter_values_start, expected)
E       AssertionError: DataFrame are different
E
E       DataFrame shape mismatch
E       [left]:  (2, 4)
E       [right]: (3, 4)
==================================== short test summary info =====================================================
FAILED pymove/tests/preprocessing/test_filters.py::test_by_datetime - AssertionError: DataFrame are different
================================= 1 failed, 94 passed, 1 warning in 3.68s ========================================

Testing new features and bugfixes

If a new the code was modified, it is essential to run against the existing tests and implement new tests if new features were created. To create a new test, you can follow the structure similar to the source code, for example

  • If creating a new function named filter_by_city in the module pymove/preprocessing/filter.py, the test must be implemented in the file pymove/tests/test_preprocessing_filters.py, named test_filter_by_city

  • If creating a new module, for example pymove/models/clustering/hierarchical.py, the tests must be placed in pymove/tests/test_models_clustering_hierarchical.py

  • Every test must be independent from another, and if using files, you must use the tmpdir directive, like bellow. More on https://code-maven.com/temporary-files-and-directory-for-pytest

def test_some_thing(tmpdir):
    d = tmpdir.mkdir('folder')

    f = d.join('name.csv')
    f.write(data)
    filename = os.path.join(
        f.dirname, f.basename
    )
    with open(filename, 'r') as f:
        ...

Coverage.py

Coverage.py allows us to se which lines of code are beign covered bt our tests To use it, install running pip install coverage or conda install -c conda-forge coverage

  • To create the report, run coverage run -m pytest

  • To view the report, run coverage report

  • The coverage html will create an html page that you can navigate, for that, open the fie htmlcov/index.html

  • To erase the data, run coverage erase