def test_fit_callable_distance(df):
clusterer = KMeansClusterer()
def dist(u, v):
return 1
clusterer.fit(df, distance=dist)
def test_predict_dataframe(df):
clusterer = KMeansClusterer()
clusterer.fit(df)
new = {'feature1': [2], 'feature2': [2]}
dataframe = pd.DataFrame(data=new)
clusterer.predict(dataframe)
def test_predict_featuresets(df):
clusterer = KMeansClusterer()
clusterer.fit(df)
new = [{'feature1': 2, 'feature2': 2}]
clusterer.predict(new)
def test_predict_dict(df):
clusterer = KMeansClusterer()
clusterer.fit(df)
new = {'feature1': 2, 'feature2': 2}
clusterer.predict(new)
def test_fit_bad_max_iter(df):
clusterer = KMeansClusterer()
with pytest.raises(ValueError):
clusterer.fit(df, max_iter=0)
def test_fit_bad_distance(df):
clusterer = KMeansClusterer()
with pytest.raises(TypeError):
clusterer.fit(df, distance=1)
def test_predict_bad_input_value(df):
clusterer = KMeansClusterer()
clusterer.fit(df)
with pytest.raises(ValueError):
clusterer.predict(pd.Series([0, 1, 2]))
def test_fit_dataframe(df):
clusterer = KMeansClusterer()
clusterer.fit(df)
def test_fit_tolerance():
data = pd.DataFrame(data={'feature': [1]})
clusterer = KMeansClusterer()
clusterer.fit(data)
def test_inertia(df):
clusterer = KMeansClusterer()
clusterer.fit(df)
assert isinstance(clusterer.inertia(), float)
def test_fit_featureset():
data = [{'feature1': 2, 'feature2': 3}, {'feature1': 7, 'feature2': 6}]
clusterer = KMeansClusterer()
clusterer.fit(data)
def test_object_creation():
KMeansClusterer()
def test_get_features(df):
clusterer = KMeansClusterer()
clusterer.fit(df)
assert clusterer.get_features() == ['feature1', 'feature2']
def test_get_clusters_not_fitted():
clusterer = KMeansClusterer()
with pytest.raises(AttributeError):
clusterer.get_clusters()
def test_predict_series(df):
clusterer = KMeansClusterer()
clusterer.fit(df)
new = {'feature1': 2, 'feature2': 2}
series = pd.Series(data=new)
clusterer.predict(series)
def test_fit_bad_data():
data = [1, 0]
clusterer = KMeansClusterer()
with pytest.raises(TypeError):
clusterer.fit(data)
def test_predict_not_fitted():
clusterer = KMeansClusterer()
new = {'feature1': 2, 'feature2': 2}
series = pd.Series(data=new)
with pytest.raises(AttributeError):
clusterer.predict(series)
def test_predict_bad_input_type(df):
clusterer = KMeansClusterer()
clusterer.fit(df)
with pytest.raises(TypeError):
clusterer.predict([0, 1])