def test_imputer_feature_boston():
o = Inputers(
description_filepath="../../descriptions/pre/inputers/boston.yaml")
train = o.transform(dataset="train")
imp = Imputers(description_filepath=
"../../descriptions/pre/cleaners/most_frequent_impute.yaml")
assert imp.transform(train, inplace=True, verbose=True).shape == (506, 14)
def test_booL_some(City):
o = Inputers(
description_filepath="../../descriptions/pre/inputers/yeast3.yaml")
train = o.transform(dataset="train")
train["Alm"] = False
train.loc[0, "Alm"] = True
assert boolean_to_integer(train, inplace=True).loc[0, "Alm"] == 1
def test_inputer_datasets():
inputer = Inputers(description_filepath="../../descriptions/pre/inputers/iris.yaml")
assert inputer.datasets() == [
"train",
"valid",
"test",
"sampleSubmission",
"directory_path",
]
def test_learn_train_kw_target_iris(flower):
inputer = Inputers(description_filepath=
"../../descriptions/pre/inputers/pima-diabetes.yaml")
dataset = inputer.transform()
y = dataset[inputer.target].values
X = dataset[dataset.columns.difference([inputer.target])]
o = Learners(
description_filepath="../../descriptions/learners/RFC.sm.yaml")
assert inputer.target == "Outcome"
def test_imputer_features_nans_found():
o = Inputers(
description_filepath="../../descriptions/pre/inputers/yeast3.yaml")
train = o.transform(dataset="train")
train["bad"] = np.NaN
imp = Imputers(description_filepath=
"../../descriptions/pre/cleaners/most_frequent_impute.yaml")
assert (imp.transform(train, features=["Alm"], inplace=True,
verbose=True).isnull().any().any()) == True
def test_learn_model_name():
inputer = Inputers(description_filepath=
"../../descriptions/pre/inputers/pima-diabetes.yaml")
dataset = inputer.transform()
y = dataset[inputer.target].values
X = dataset[dataset.columns.difference([inputer.target])]
learner = Learners(
description_filepath="../../descriptions/learners/RFC.yaml")
learner.train(X, y, checkpoint="RandomForest.ckp")
assert learner.model_name == "RandomForest"
def test_inputer_transform_splitter_X_valid():
inputer = Inputers(description_filepath="../../descriptions/pre/inputers/iris.yaml")
train = inputer.transform()
y = train[inputer.target].values
X = train[train.columns.difference([inputer.target])]
splitter = Splitters(
description_filepath="../../descriptions/pre/inputers/split_30_stratify.yaml"
)
train, valid, y_train, y_valid = splitter.transform(X, y)
assert valid.shape == (45, 4) and y_valid.shape == (45,)
def test_evaluate_otto_group():
inputer = Inputers(
description_filepath="../../descriptions/pre/inputers/otto_group.yaml")
dataset = inputer.transform()
y = dataset[inputer.target].values
X = dataset[dataset.columns.difference([inputer.target])]
learner = Learners(
description_filepath="../../descriptions/learners/XGBC.yaml")
learner.train(X, y, checkpoint="pima_LGBMC.ckp")
assert len(learner.evaluate(X, y).keys()) == 7
def test_train_predict_prob_LGBC_wine():
inputer = Inputers(
description_filepath="../../descriptions/pre/inputers/wine.yaml")
wine = inputer.transform()
y = wine[inputer.target].values
X = wine[wine.columns.difference([inputer.target])]
learner = Learners(
description_filepath="../../descriptions/learners/LGBC.yaml")
learner.train(X, y, target=inputer.target, checkpoint="wineLGBMC.ckp")
assert learner.predict_proba(X).shape == (178, 3)
def test_train_predict_prob_XGBC_yeast3():
inputer = Inputers(
description_filepath="../../descriptions/pre/inputers/yeast3.yaml")
yeast3 = inputer.transform()
y = yeast3[inputer.target].values
X = yeast3[yeast3.columns.difference([inputer.target])]
learner = Learners(
description_filepath="../../descriptions/learners/XGBC.yaml")
learner.train(X, y, checkpoint="yeast3LGBMC.ckp")
assert learner.predict_proba(X).shape == (1484, 2)
def test_train_predict_prob_LGBC():
inputer = Inputers(description_filepath=
"../../descriptions/pre/inputers/pima-diabetes.yaml")
diabetes = inputer.transform()
y = diabetes[inputer.target].values
X = diabetes[diabetes.columns.difference([inputer.target])]
learner = Learners(
description_filepath="../../descriptions/learners/LGBC.yaml")
learner.train(X, y, checkpoint="diabetesLGBMC.ckp")
assert learner.predict_proba(X).shape == (768, 2)
def test_learn_train_predict_prob_LGBC():
inputer = Inputers(
description_filepath="../../descriptions/pre/inputers/otto_group.yaml")
dataset = inputer.transform()
y = dataset[inputer.target].values
X = dataset[dataset.columns.difference([inputer.target])]
learner = Learners(
description_filepath="../../descriptions/learners/LGBC.yaml")
learner.train(X, y, checkpoint="otto_group_LGBMC.ckp")
assert learner.predict(X).shape == (61878, )
def test_learn_train_predict_XGBC():
inputer = Inputers(
description_filepath="../../descriptions/pre/inputers/wine.yaml")
dataset = inputer.transform()
y = dataset[inputer.target].values
X = dataset[dataset.columns.difference([inputer.target])]
learner = Learners(
description_filepath="../../descriptions/learners/XGBC.yaml")
learner.train(X, y, checkpoint="LGBMClassifier.ckp")
assert learner.predict(X).shape == (178, )
def test_imputer_feature_boston_err():
o = Inputers(
description_filepath="../../descriptions/pre/inputers/boston.yaml")
train = o.transform(dataset="train")
train["ggoo"] = np.nan
imp = Imputers(description_filepath=
"../../descriptions/pre/cleaners/most_frequent_impute.yaml")
with pytest.raises(PasoError):
assert imp.transform(train, inplace=True,
verbose=True).shape == (506, 14)
def test_learn_train_kw_target_pima():
inputer = Inputers(
description_filepath="../../descriptions/pre/inputers/wine.yaml")
dataset = inputer.transform()
y = dataset[inputer.target].values
X = dataset[dataset.columns.difference([inputer.target])]
learner = Learners(
description_filepath="../../descriptions/learners/RFC.yaml")
learner.train(X, y, checkpoint="diabetesRandomForest1.ckp")
assert learner.trained == True
def test_train_predict_prob_XGBC_creditdard():
inputer = Inputers(
description_filepath="../../descriptions/pre/inputers/creditcard.yaml")
creditdard = inputer.transform()
y = creditdard[inputer.target].values
X = creditdard[creditdard.columns.difference([inputer.target])]
learner = Learners(
description_filepath="../../descriptions/learners/XGBC.yaml")
learner.train(X, y, checkpoint="creditdardLGBMC.ckp")
assert learner.predict_proba(X).shape == (284807, 2)
def test_learn_train_no_y():
inputer = Inputers(description_filepath=
"../../descriptions/pre/inputers/pima-diabetes.yaml")
dataset = inputer.transform()
y = dataset[inputer.target].values
X = dataset[dataset.columns.difference([inputer.target])]
o = Learners(
description_filepath="../../descriptions/learners/RFC.sm.yaml")
with pytest.raises(IndexError):
o.train(X) == o
def test_inputer_transform_splitter_onto_wrong_place():
inputer = Inputers(description_filepath="../../descriptions/pre/inputers/iris.yaml")
Flower = inputer.transform()
splitter = Splitters()
with pytest.raises(PasoError):
train, valid = splitter.transform(
Flower,
target=inputer.target,
description_filepath="../../descriptions/pre/inputers/test_size_30.yaml",
)
def test_imputer_features_not_found():
o = Inputers(
description_filepath="../../descriptions/pre/inputers/yeast3.yaml")
train = o.transform(dataset="train")
imp = Imputers(description_filepath=
"../../descriptions/pre/cleaners/most_frequent_impute.yaml")
with pytest.raises(PasoError):
train = imp.transform(train,
features=["bad", "badder", "Alm"],
inplace=True,
verbose=True)
def test_inputer_imputer():
inputer = Inputers(description_filepath="../../descriptions/pre/inputers/iris.yaml")
assert inputer.inputers() == [
"exec",
"cvs",
"xls",
"xlsm",
"text",
"image2D",
"image3D",
]
def test_spltter_transform_creditcard_20__url_cvs_zip():
inputer = Inputers(
description_filepath="../../descriptions/pre/inputers/creditcard.yaml"
)
train = inputer.transform(dataset="train")
y = train[inputer.target].values
X = train[train.columns.difference([inputer.target])]
splitter = Splitters(
description_filepath="../../descriptions/pre/inputers/split_20_stratify.yaml"
)
train, valid, _, _ = splitter.transform(X, y)
assert train.shape == (227845, 30) and valid.shape == (56962, 30)
def test_predict_Prob_error():
inputer = Inputers(description_filepath=
"../../descriptions/pre/inputers/pima-diabetes.yaml")
diabetes = inputer.transform()
learner = Learners(
description_filepath="../../descriptions/learners/LGBC.yaml")
# learner.train(
# diabetes, target=inputer.target, checkpoint="diabetesLGBMC.ckp"
# )
X = diabetes
X_train = X[X.columns.difference([inputer.target])]
with pytest.raises(PasoError):
assert learner.predict_proba(X_train).shape == (768, 2)
def test_imputer_features_allnans():
o = Inputers(
description_filepath="../../descriptions/pre/inputers/yeast3.yaml")
train = o.transform(dataset="train")
train["Alm"] = np.NaN
train.loc[0, "Alm"] = 0
imp = Imputers(description_filepath=
"../../descriptions/pre/cleaners/most_frequent_impute.yaml")
with pytest.raises(PasoError):
assert (imp.transform(train,
features=["Alm"],
inplace=True,
verbose=True).isnull().all().all()) == False
def test_splitter_transform__onto_group():
inputer = Inputers(description_filepath="../../descriptions/pre/inputers/iris.yaml")
train = inputer.transform()
y = train[inputer.target].values
X = train[train.columns.difference([inputer.target])]
splitter = Splitters(
description_filepath="../../descriptions/pre/inputers/split_20_stratify.yaml"
)
train, valid, y_train, y_valid = splitter.transform(X, y)
assert (
train.shape == (120, 4)
and valid.shape == (30, 4)
and y_train.shape == (120,)
and y_valid.shape == (30,)
)
def test_spitter_transform_s_wine():
inputer = Inputers(description_filepath="../../descriptions/pre/inputers/wine.yaml")
train = inputer.transform()
y = train[inputer.target].values
X = train[train.columns.difference([inputer.target])]
splitter = Splitters(
description_filepath="../../descriptions/pre/inputers/split_30_stratify.yaml"
)
train, valid, y_train, y_valid = splitter.transform(X, y)
assert (
train.shape == (124, 13)
and valid.shape == (54, 13)
and y_train.shape == (124,)
and y_valid.shape == (54,)
)
def test_learner_cross_validate_LGBC():
inputer = Inputers(description_filepath=
"../../descriptions/pre/inputers/pima-diabetes.yaml")
dataset = inputer.transform()
y = dataset[inputer.target].values
X = dataset[dataset.columns.difference([inputer.target])]
learner = Learners(
description_filepath="../../descriptions/learners/LGBC.yaml")
learner.train(X, y, checkpoint="pima_LGBMC.ckp")
learner.cross_validate(
X,
y,
cv_description_filepath=
"../../descriptions/learners/Cross_validation_classification.yaml",
)
assert len(learner.evaluate(X, y).keys()) == 9
def test_learner_cross_validate_RFC_iris_multiclass_evaluate_AO():
inputer = Inputers(
description_filepath="../../descriptions/pre/inputers/iris.yaml")
dataset = inputer.transform()
y = dataset[inputer.target].values
X = dataset[dataset.columns.difference([inputer.target])]
learner = Learners(
description_filepath="../../descriptions/learners/RFC.yaml")
learner.train(X, y, checkpoint="iris)RC.ckp")
learner.cross_validate(
X,
y,
cv_description_filepath=
"../../descriptions/learners/Cross_validation_classification.yaml",
)
assert learner.evaluate(X, y)["accuracy"] == 1.0
def test_spltter_transform_creditcard_url_30__cvs_zip():
inputer = Inputers(
description_filepath="../../descriptions/pre/inputers/yeast3.yaml"
)
train = inputer.transform(dataset="train")
y = train[inputer.target].values
X = train[train.columns.difference([inputer.target])]
splitter = Splitters(
description_filepath="../../descriptions/pre/inputers/split_20_stratify.yaml"
)
train, valid, y_train, y_valid = splitter.transform(X, y)
assert (
train.shape == (1187, 8)
and valid.shape == (297, 8)
and y_train.shape == (1187,)
and y_valid.shape == (297,)
)
def test_learner_cross_validate_RFC_iris_milticlass_evaluate_test_accuracy():
inputer = Inputers(
description_filepath="../../descriptions/pre/inputers/iris.yaml")
dataset = inputer.transform()
y = dataset[inputer.target].values
X = dataset[dataset.columns.difference([inputer.target])]
learner = Learners(
description_filepath="../../descriptions/learners/RFC.yaml")
learner.train(X, y, checkpoint="pima_LGBMC.ckp")
score = learner.cross_validate(
X,
y,
cv_description_filepath=
"../../descriptions/learners/Cross_validation_classification.yaml",
)
assert score["mean"]["test_accuracy"] >= 0.95
def test_inputer_transform_cvs_url():
link = "https://archive.ics.uci.edu/ml/machine-learning-databases/wine/wine.data"
names = [
"Class",
"Alcohol",
"Malic-acid",
"Ash",
"Alcalinity-ash",
"Magnesium",
"phenols",
"Flavanoids",
"Nonflavanoid-phenols",
"Proanthocyanins",
"Color-intensity",
"Hue",
"OD280-OD315-diluted-wines",
"Proline",
]
winmeo = pd.read_csv(link, names=names).head()
inputer = Inputers(description_filepath="../../descriptions/pre/inputers/wine.yaml")
assert (inputer.transform().columns == winmeo.columns).any()