def test_date_encoder_returns_only_day(self, dates: pd.DataFrame):
date_coder = DateEncoder(day=True, month=False, week=False, year=False)
result = date_coder.fit_transform(dates)
assert isinstance(result, pd.DataFrame)
assert 1 == len(result.columns)
assert len(dates) == len(result)
assert "date_a_year" not in result.columns
assert "date_a_day" in result.columns
assert "date_a_month" not in result.columns
assert "date_a_week" not in result.columns
def test_date_encoder_returns_only_week(dates):
week_coder = DateEncoder(day=False, month=False, week=True, year=False)
result = week_coder.fit_transform(dates)
assert isinstance(result, pd.DataFrame)
assert 1 == len(result.columns)
assert len(dates) == len(result)
assert 'date_a_year' not in result.columns
assert 'date_a_day' not in result.columns
assert 'date_a_month' not in result.columns
assert 'date_a_week' in result.columns
def test_date_encoder_returns_correctly(dates):
date_coder = DateEncoder()
result = date_coder.fit_transform(dates)
assert isinstance(result, pd.DataFrame)
assert 4 == len(result.columns)
assert len(dates) == len(result)
for col in result.columns:
assert pd.api.types.is_numeric_dtype(result[col])
assert 'date_a' not in result.columns
assert 'date_a_year' in result.columns
assert 'date_a_day' in result.columns
assert 'date_a_month' in result.columns
assert 'date_a_week' in result.columns
def test_date_encoder_works_in_grid_search(self, dates: pd.DataFrame):
pipe = create_pipeline(DateEncoder())
grid = GridSearchCV(
pipe,
param_grid={"clf__strategy": ["stratified", "most_frequent"]},
cv=2,
)
grid.fit(dates, [0, 0, 1, 1])
assert hasattr(grid, "best_score_")
def test_date_encoder_works_in_cv(self, dates: pd.DataFrame):
pipe = create_pipeline(DateEncoder())
score = cross_val_score(pipe, dates, y=[0, 0, 1, 1], n_jobs=2, cv=2)
assert 2 == len(score)
def test_works_without_args(self):
assert DateEncoder()
"""
host_since
==========
When started hosting. Hypothesis that being a host for longer affects the price - they might be able to charge a different price.
For our solution, we can set it to 0 or ask.
Is a date - note that date is not a dtype, but we can set read_csv to parse it automatically as a date
dtype: datetime
"""
from ml_tooling.transformers import Select, DateEncoder
from sklearn.pipeline import Pipeline
host_since = Pipeline([("select", Select("host_since")),
("date_encoder", DateEncoder())])