spaCy/spacy/tests
Raphael Mitsch c0fd8a2e71
find-threshold: CLI command for multi-label classifier threshold tuning (#11280)
* Add foundation for find-threshold CLI functionality.

* Finish first draft for find-threshold.

* Add tests.

* Revert adjusted import statements.

* Fix mypy errors.

* Fix imports.

* Harmonize arguments with spacy evaluate command.

* Generalize component and threshold handling. Harmonize arguments with 'spacy evaluate' CLI.

* Fix Spancat test.

* Add beta parameter to Scorer and PRFScore.

* Make beta a component scorer setting.

* Remove beta.

* Update nlp.config (workaround).

* Reload pipeline on threshold change. Adjust tests. Remove confection reference.

* Remove assumption of component being a Pipe object or having a .cfg attribute.

* Adjust test output and reference values.

* Remove beta references. Delete universe.json.

* Reverting unnecessary changes. Removing unused default values. Renaming variables in find-cli tests.

* Update spacy/cli/find_threshold.py

Co-authored-by: Adriane Boyd <adrianeboyd@gmail.com>

* Remove adding labels in tests.

* Remove unused error

* Undo changes to PRFScorer

* Change default value for n_trials. Log table iteratively.

* Add warnings for pointless applications of find_threshold().

* Fix imports.

* Adjust type check of TextCategorizer to exclude subclasses.

* Change check of if there's only one unique value in scores.

* Update spacy/cli/find_threshold.py

Co-authored-by: Sofie Van Landeghem <svlandeg@users.noreply.github.com>

* Incorporate feedback.

* Fix test issue. Update docstring.

* Update docs & docstring.

* Update spacy/tests/test_cli.py

Co-authored-by: Adriane Boyd <adrianeboyd@gmail.com>

* Add examples to docs. Rename _nlp to nlp in tests.

* Update spacy/cli/find_threshold.py

Co-authored-by: Sofie Van Landeghem <svlandeg@users.noreply.github.com>

* Update spacy/cli/find_threshold.py

Co-authored-by: Sofie Van Landeghem <svlandeg@users.noreply.github.com>

Co-authored-by: Adriane Boyd <adrianeboyd@gmail.com>
Co-authored-by: Sofie Van Landeghem <svlandeg@users.noreply.github.com>
2022-11-25 11:44:55 +01:00
..
doc Restore Doc attr getter values in Doc.to_json (#11700) 2022-11-03 11:49:08 +01:00
lang Update Russian and Ukrainian lemmatizers (#11811) 2022-11-25 11:12:46 +01:00
matcher Auto-format code with black (#11513) 2022-09-16 11:50:19 +02:00
morphology
package Add a way to get the URL to download a pipeline to the CLI (#11175) 2022-09-02 11:58:21 +02:00
parser Auto-format code with black (#11022) 2022-06-27 09:35:35 +02:00
pipeline Check textcat values for validity (#11763) 2022-11-17 10:25:01 +01:00
serialize Revert disable/disabled merging behavior (#11745) 2022-11-08 14:58:10 +01:00
tokenizer Fuzz tokenizer.explain: draft for fuzzy tests. (#10771) 2022-05-17 10:23:16 +02:00
training Add `training.before_update` callback (#11739) 2022-11-23 17:54:58 +01:00
vocab_vectors fix comparison of constants (#11834) 2022-11-21 08:12:03 +01:00
README.md Update docs [ci skip] 2020-11-09 12:43:26 +08:00
__init__.py
conftest.py Update Russian and Ukrainian lemmatizers (#11811) 2022-11-25 11:12:46 +01:00
enable_gpu.py Set up GPU CI testing (#7293) 2021-04-22 14:58:29 +02:00
test_architectures.py Tidy up code 2021-06-28 12:08:15 +02:00
test_cli.py find-threshold: CLI command for multi-label classifier threshold tuning (#11280) 2022-11-25 11:44:55 +01:00
test_displacy.py #10672: fixes displacy output for manual unsorted entities (#10673) 2022-04-27 09:51:58 +02:00
test_errors.py use metaclass to decorate errors (#9593) 2021-11-03 15:29:32 +01:00
test_language.py Check for any non-Doc returned value for components (#11424) 2022-09-01 19:37:23 +02:00
test_misc.py Fix `test_{prefer,require}_gpu` (#11390) 2022-08-30 14:21:02 +02:00
test_models.py Rename test helper method with non-test_ name (#11701) 2022-10-25 14:53:18 +02:00
test_pickles.py Include noun chunks method when pickling Vocab 2021-02-12 13:27:46 +01:00
test_scorer.py Update textcat scorer threshold behavior (#11696) 2022-11-02 15:35:04 +01:00
test_ty.py Custom component types in spacy.ty (#9469) 2021-10-21 15:31:06 +02:00
util.py account for NER labels with a hyphen in the name (#10960) 2022-06-17 20:02:37 +01:00

README.md

spaCy tests

spaCy uses the pytest framework for testing. For more info on this, see the pytest documentation.

Tests for spaCy modules and classes live in their own directories of the same name. For example, tests for the Tokenizer can be found in /tests/tokenizer. All test modules (i.e. directories) also need to be listed in spaCy's setup.py. To be interpreted and run, all test files and test functions need to be prefixed with test_.

⚠️ Important note: As part of our new model training infrastructure, we've moved all model tests to the spacy-models repository. This allows us to test the models separately from the core library functionality.

Table of contents

  1. Running the tests
  2. Dos and don'ts
  3. Parameters
  4. Fixtures
  5. Helpers and utilities
  6. Contributing to the tests

Running the tests

To show print statements, run the tests with py.test -s. To abort after the first failure, run them with py.test -x.

py.test spacy                        # run basic tests
py.test spacy --slow                 # run basic and slow tests

You can also run tests in a specific file or directory, or even only one specific test:

py.test spacy/tests/tokenizer  # run all tests in directory
py.test spacy/tests/tokenizer/test_exceptions.py # run all tests in file
py.test spacy/tests/tokenizer/test_exceptions.py::test_tokenizer_handles_emoji # run specific test

Dos and don'ts

To keep the behavior of the tests consistent and predictable, we try to follow a few basic conventions:

  • Test names should follow a pattern of test_[module]_[tested behaviour]. For example: test_tokenizer_keeps_email or test_spans_override_sentiment.
  • If you're testing for a bug reported in a specific issue, always create a regression test. Regression tests should be named test_issue[ISSUE NUMBER] and live in the regression directory.
  • Only use @pytest.mark.xfail for tests that should pass, but currently fail. To test for desired negative behavior, use assert not in your test.
  • Very extensive tests that take a long time to run should be marked with @pytest.mark.slow. If your slow test is testing important behavior, consider adding an additional simpler version.
  • If tests require loading the models, they should be added to the spacy-models tests.
  • Before requiring the models, always make sure there is no other way to test the particular behavior. In a lot of cases, it's sufficient to simply create a Doc object manually. See the section on helpers and utility functions for more info on this.
  • Avoid unnecessary imports. There should never be a need to explicitly import spaCy at the top of a file, and many components are available as fixtures. You should also avoid wildcard imports (from module import *).
  • If you're importing from spaCy, always use absolute imports. For example: from spacy.language import Language.
  • Try to keep the tests readable and concise. Use clear and descriptive variable names (doc, tokens and text are great), keep it short and only test for one behavior at a time.

Parameters

If the test cases can be extracted from the test, always parametrize them instead of hard-coding them into the test:

@pytest.mark.parametrize('text', ["google.com", "spacy.io"])
def test_tokenizer_keep_urls(tokenizer, text):
    tokens = tokenizer(text)
    assert len(tokens) == 1

This will run the test once for each text value. Even if you're only testing one example, it's usually best to specify it as a parameter. This will later make it easier for others to quickly add additional test cases without having to modify the test.

You can also specify parameters as tuples to test with multiple values per test:

@pytest.mark.parametrize('text,length', [("U.S.", 1), ("us.", 2), ("(U.S.", 2)])

To test for combinations of parameters, you can add several parametrize markers:

@pytest.mark.parametrize('text', ["A test sentence", "Another sentence"])
@pytest.mark.parametrize('punct', ['.', '!', '?'])

This will run the test with all combinations of the two parameters text and punct. Use this feature sparingly, though, as it can easily cause unnecessary or undesired test bloat.

Fixtures

Fixtures to create instances of spaCy objects and other components should only be defined once in the global conftest.py. We avoid having per-directory conftest files, as this can easily lead to confusion.

These are the main fixtures that are currently available:

Fixture Description
tokenizer Basic, language-independent tokenizer. Identical to the xx language class.
en_tokenizer, de_tokenizer, ... Creates an English, German etc. tokenizer.
en_vocab Creates an instance of the English Vocab.

The fixtures can be used in all tests by simply setting them as an argument, like this:

def test_module_do_something(en_tokenizer):
    tokens = en_tokenizer("Some text here")

If all tests in a file require a specific configuration, or use the same complex example, it can be helpful to create a separate fixture. This fixture should be added at the top of each file. Make sure to use descriptive names for these fixtures and don't override any of the global fixtures listed above. From looking at a test, it should immediately be clear which fixtures are used, and where they are coming from.

Helpers and utilities

Our new test setup comes with a few handy utility functions that can be imported from util.py.

Constructing a Doc object manually

Loading the models is expensive and not necessary if you're not actually testing the model performance. If all you need is a Doc object with annotations like heads, POS tags or the dependency parse, you can construct it manually.

def test_doc_token_api_strings(en_vocab):
    words = ["Give", "it", "back", "!", "He", "pleaded", "."]
    pos = ['VERB', 'PRON', 'PART', 'PUNCT', 'PRON', 'VERB', 'PUNCT']
    heads = [0, 0, 0, 0, 5, 5, 5]
    deps = ['ROOT', 'dobj', 'prt', 'punct', 'nsubj', 'ROOT', 'punct']

    doc = Doc(en_vocab, words=words, pos=pos, heads=heads, deps=deps)
    assert doc[0].text == 'Give'
    assert doc[0].lower_ == 'give'
    assert doc[0].pos_ == 'VERB'
    assert doc[0].dep_ == 'ROOT'

Other utilities

Name Description
apply_transition_sequence(parser, doc, sequence) Perform a series of pre-specified transitions, to put the parser in a desired state.
add_vecs_to_vocab(vocab, vectors) Add list of vector tuples ([("text", [1, 2, 3])]) to given vocab. All vectors need to have the same length.
get_cosine(vec1, vec2) Get cosine for two given vectors.
assert_docs_equal(doc1, doc2) Compare two Doc objects and assert that they're equal. Tests for tokens, tags, dependencies and entities.

Contributing to the tests

There's still a long way to go to finally reach 100% test coverage and we'd appreciate your help! 🙌 You can open an issue on our issue tracker and label it tests, or make a pull request to this repository.

📖 For more information on contributing to spaCy in general, check out our contribution guidelines.