Thanks to visit codestin.com
Credit goes to github.com

Skip to content

Commit bb6117b

Browse files
thomasjpfanthomasjpfan
authored
TST Enable numpydoc validation default (#20304)
1 parent 617ff6e commit bb6117b

File tree

3 files changed

+231
-44
lines changed

3 files changed

+231
-44
lines changed

‎build_tools/azure/posix.yml‎

Lines changed: 4 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -58,6 +58,10 @@ jobs:
5858
- script: |
5959
build_tools/azure/test_docs.sh
6060
displayName: 'Test Docs'
61+
- script: |
62+
build_tools/azure/test_docstring.sh
63+
displayName: "Numpydoc validation"
64+
condition: eq(variables['TEST_DOCSTRINGS'], 'true')
6165
- script: |
6266
build_tools/azure/test_pytest_soft_dependency.sh
6367
displayName: 'Test Soft Dependency'

‎build_tools/azure/test_docstring.sh‎

Lines changed: 15 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,15 @@
1+
#!/bin/bash
2+
3+
set -e
4+
5+
if [[ "$DISTRIB" =~ ^conda.* ]]; then
6+
source activate $VIRTUALENV
7+
elif [[ "$DISTRIB" == "ubuntu" ]]; then
8+
source $VIRTUALENV/bin/activate
9+
fi
10+
11+
if [[ "$BUILD_WITH_ICC" == "true" ]]; then
12+
source /opt/intel/oneapi/setvars.sh
13+
fi
14+
15+
pytest maint_tools/test_docstrings.py

‎maint_tools/test_docstrings.py‎

Lines changed: 212 additions & 44 deletions
Original file line numberDiff line numberDiff line change
@@ -3,53 +3,214 @@
33
from typing import Optional
44

55
import pytest
6-
from sklearn.utils._testing import all_estimators
6+
from sklearn.utils import all_estimators
77

88
numpydoc_validation = pytest.importorskip("numpydoc.validate")
99

10-
# List of whitelisted modules and methods; regexp are supported.
11-
DOCSTRING_WHITELIST = [
12-
"LogisticRegression$",
13-
"LogisticRegression.fit",
14-
"LogisticRegression.decision_function",
15-
"Birch.predict",
16-
"Birch.transform",
10+
# List of modules ignored when checking for numpydoc validation.
11+
DOCSTRING_IGNORE_LIST = [
12+
"ARDRegression",
13+
"AdaBoostClassifier",
14+
"AdaBoostRegressor",
15+
"AdditiveChi2Sampler",
16+
"AffinityPropagation",
17+
"AgglomerativeClustering",
18+
"BaggingClassifier",
19+
"BaggingRegressor",
20+
"BayesianGaussianMixture",
21+
"BayesianRidge",
22+
"BernoulliNB",
23+
"BernoulliRBM",
24+
"Binarizer",
25+
"Birch",
26+
"CCA",
27+
"CalibratedClassifierCV",
28+
"CategoricalNB",
29+
"ClassifierChain",
30+
"ColumnTransformer",
31+
"ComplementNB",
32+
"CountVectorizer",
33+
"DBSCAN",
34+
"DecisionTreeClassifier",
35+
"DecisionTreeRegressor",
36+
"DictVectorizer",
37+
"DictionaryLearning",
38+
"DummyClassifier",
39+
"DummyRegressor",
40+
"ElasticNet",
41+
"ElasticNetCV",
42+
"EllipticEnvelope",
43+
"EmpiricalCovariance",
44+
"ExtraTreeClassifier",
45+
"ExtraTreeRegressor",
46+
"ExtraTreesClassifier",
47+
"ExtraTreesRegressor",
48+
"FactorAnalysis",
49+
"FastICA",
50+
"FeatureAgglomeration",
51+
"FeatureHasher",
52+
"FeatureUnion",
53+
"FunctionTransformer",
54+
"GammaRegressor",
55+
"GaussianMixture",
56+
"GaussianNB",
57+
"GaussianProcessClassifier",
58+
"GaussianProcessRegressor",
59+
"GaussianRandomProjection",
60+
"GenericUnivariateSelect",
1761
"GradientBoostingClassifier",
1862
"GradientBoostingRegressor",
19-
"LinearDiscriminantAnalysis.decision_function",
20-
"LinearSVC.decision_function",
21-
"LogisticRegressionCV.decision_function",
22-
"OPTICS",
23-
"OPTICS.fit",
24-
"PassiveAggressiveClassifier.decision_function",
25-
"Perceptron.decision_function",
26-
"RidgeClassifier.decision_function",
27-
"RidgeClassifier.fit",
28-
"RidgeClassifierCV.decision_function",
63+
"GraphicalLasso",
64+
"GraphicalLassoCV",
65+
"GridSearchCV",
66+
"HalvingGridSearchCV",
67+
"HalvingRandomSearchCV",
68+
"HashingVectorizer",
69+
"HistGradientBoostingClassifier",
70+
"HistGradientBoostingRegressor",
71+
"HuberRegressor",
72+
"IncrementalPCA",
73+
"IsolationForest",
74+
"Isomap",
75+
"IsotonicRegression",
76+
"IterativeImputer",
77+
"KBinsDiscretizer",
78+
"KMeans",
79+
"KNNImputer",
80+
"KNeighborsClassifier",
81+
"KNeighborsRegressor",
82+
"KNeighborsTransformer",
83+
"KernelCenterer",
2984
"KernelDensity",
30-
"KernelDensity.fit",
31-
"KernelDensity.score",
32-
"DecisionTreeClassifier",
33-
"DecisionTreeRegressor",
34-
"LinearRegression$",
35-
"SGDClassifier.decision_function",
36-
"SGDClassifier.set_params",
37-
"SGDClassifier.get_params",
38-
"SGDClassifier.fit",
39-
"SGDClassifier.partial_fit",
40-
"SGDClassifier.predict$", # $ to avoid match w/ predict_proba (regex)
41-
"SGDClassifier.score",
42-
"SGDClassifier.sparsify",
43-
"SGDClassifier.densify",
44-
"VotingClassifier.fit",
45-
"VotingClassifier.transform",
46-
"VotingClassifier.predict",
47-
"VotingClassifier.score",
48-
"VotingClassifier.predict_proba",
49-
"VotingClassifier.set_params",
50-
"VotingClassifier.get_params",
51-
"VotingClassifier.named_estimators",
52-
"VotingClassifier$",
85+
"KernelPCA",
86+
"KernelRidge",
87+
"LabelBinarizer",
88+
"LabelEncoder",
89+
"LabelPropagation",
90+
"LabelSpreading",
91+
"Lars",
92+
"LarsCV",
93+
"Lasso",
94+
"LassoCV",
95+
"LassoLars",
96+
"LassoLarsCV",
97+
"LassoLarsIC",
98+
"LatentDirichletAllocation",
99+
"LedoitWolf",
100+
"LinearDiscriminantAnalysis",
101+
"LinearRegression",
102+
"LinearSVC",
103+
"LinearSVR",
104+
"LocalOutlierFactor",
105+
"LocallyLinearEmbedding",
106+
"LogisticRegression",
107+
"LogisticRegressionCV",
108+
"MDS",
109+
"MLPClassifier",
110+
"MLPRegressor",
111+
"MaxAbsScaler",
112+
"MeanShift",
113+
"MinCovDet",
114+
"MinMaxScaler",
115+
"MiniBatchDictionaryLearning",
116+
"MiniBatchKMeans",
117+
"MiniBatchSparsePCA",
118+
"MissingIndicator",
119+
"MultiLabelBinarizer",
120+
"MultiOutputClassifier",
121+
"MultiOutputRegressor",
122+
"MultiTaskElasticNet",
123+
"MultiTaskElasticNetCV",
124+
"MultiTaskLasso",
125+
"MultiTaskLassoCV",
126+
"MultinomialNB",
127+
"NMF",
128+
"NearestCentroid",
129+
"NearestNeighbors",
130+
"NeighborhoodComponentsAnalysis",
131+
"Normalizer",
132+
"NuSVC",
133+
"NuSVR",
134+
"Nystroem",
135+
"OAS",
136+
"OPTICS",
137+
"OneClassSVM",
138+
"OneHotEncoder",
139+
"OneVsOneClassifier",
140+
"OneVsRestClassifier",
141+
"OrdinalEncoder",
142+
"OrthogonalMatchingPursuit",
143+
"OrthogonalMatchingPursuitCV",
144+
"OutputCodeClassifier",
145+
"PCA",
146+
"PLSCanonical",
147+
"PLSRegression",
148+
"PLSSVD",
149+
"PassiveAggressiveClassifier",
150+
"PassiveAggressiveRegressor",
151+
"PatchExtractor",
152+
"Perceptron",
153+
"Pipeline",
154+
"PoissonRegressor",
155+
"PolynomialCountSketch",
156+
"PolynomialFeatures",
157+
"PowerTransformer",
158+
"QuadraticDiscriminantAnalysis",
159+
"QuantileRegressor",
160+
"QuantileTransformer",
161+
"RANSACRegressor",
162+
"RBFSampler",
163+
"RFE",
164+
"RFECV",
165+
"RadiusNeighborsClassifier",
166+
"RadiusNeighborsRegressor",
167+
"RadiusNeighborsTransformer",
168+
"RandomForestClassifier",
169+
"RandomForestRegressor",
170+
"RandomTreesEmbedding",
171+
"RandomizedSearchCV",
172+
"RegressorChain",
173+
"Ridge",
174+
"RidgeCV",
175+
"RidgeClassifier",
176+
"RidgeClassifierCV",
177+
"RobustScaler",
178+
"SGDOneClassSVM",
179+
"SGDRegressor",
180+
"SVC",
181+
"SVR",
182+
"SelectFdr",
183+
"SelectFpr",
184+
"SelectFromModel",
185+
"SelectFwe",
186+
"SelectKBest",
187+
"SelectPercentile",
188+
"SelfTrainingClassifier",
189+
"SequentialFeatureSelector",
190+
"ShrunkCovariance",
191+
"SimpleImputer",
192+
"SkewedChi2Sampler",
193+
"SparseCoder",
194+
"SparsePCA",
195+
"SparseRandomProjection",
196+
"SpectralBiclustering",
197+
"SpectralClustering",
198+
"SpectralCoclustering",
199+
"SpectralEmbedding",
200+
"SplineTransformer",
201+
"StackingClassifier",
202+
"StackingRegressor",
203+
"StandardScaler",
204+
"TSNE",
205+
"TfidfTransformer",
206+
"TfidfVectorizer",
207+
"TheilSenRegressor",
208+
"TransformedTargetRegressor",
209+
"TruncatedSVD",
210+
"TweedieRegressor",
211+
"VarianceThreshold",
212+
"VotingClassifier",
213+
"VotingRegressor",
53214
]
54215

55216

@@ -72,7 +233,7 @@ def get_all_methods():
72233
yield Estimator, method
73234

74235

75-
def filter_errors(errors, method):
236+
def filter_errors(errors, method, Estimator=None):
76237
"""
77238
Ignore some errors based on the method type.
78239
@@ -90,6 +251,13 @@ def filter_errors(errors, method):
90251
if code in ["RT02", "GL01"]:
91252
continue
92253

254+
# Ignore PR02: Unknown parameters for properties. We sometimes use
255+
# properties for ducktyping, i.e. SGDClassifier.predict_proba
256+
if code == "PR02" and Estimator is not None and method is not None:
257+
method_obj = getattr(Estimator, method)
258+
if isinstance(method_obj, property):
259+
continue
260+
93261
# Following codes are only taken into account for the
94262
# top level class docstrings:
95263
# - ES01: No extended summary found
@@ -165,14 +333,14 @@ def test_docstring(Estimator, method, request):
165333

166334
import_path = ".".join(import_path)
167335

168-
if not any(re.search(regex, import_path) for regex in DOCSTRING_WHITELIST):
336+
if any(re.search(regex, import_path) for regex in DOCSTRING_IGNORE_LIST):
169337
request.applymarker(
170338
pytest.mark.xfail(run=False, reason="TODO pass numpydoc validation")
171339
)
172340

173341
res = numpydoc_validation.validate(import_path)
174342

175-
res["errors"] = list(filter_errors(res["errors"], method))
343+
res["errors"] = list(filter_errors(res["errors"], method, Estimator=Estimator))
176344

177345
if res["errors"]:
178346
msg = repr_errors(res, Estimator, method)

0 commit comments

Comments
 (0)