Skip to content

Add option day_number_in_year to DateFlagsTransform #552

Merged
merged 9 commits into from
Feb 22, 2022
Merged

Add option day_number_in_year to DateFlagsTransform #552

Show file tree
Hide file tree
Changes from 3 commits
Commits
Show all changes
9 commits
Select commit Hold shift + click to select a range
d140582
Add day_number_in_year option in DateFlagsTransform and tests for it
Feb 21, 2022
bd798bb
Merge remote-tracking branch 'origin/master' into issue-546
Feb 21, 2022
823aa10
Update changelog
Feb 21, 2022
1b9ad45
Fix parameter description
Feb 21, 2022
3d917cc
Fix generation of regressor prefix
Feb 21, 2022
dc912b3
Temp fix for binder
Feb 21, 2022
8eb82ce
Merge remote-tracking branch 'origin/master' into issue-546
Feb 21, 2022
0bc21b3
Fix binder back, add notebook with fixed DateFlag representation
Feb 21, 2022
06b7e35
Merge branch 'master' into issue-546
alex-hse-repository Feb 22, 2022
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion CHANGELOG.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -16,7 +16,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
- Add plot_time_series_with_change_points function ([#534](https://github.com/tinkoff-ai/etna/pull/534))
-
- Add find_change_points function ([#521](https://github.com/tinkoff-ai/etna/pull/521))
-
- Add option day_number_in_year to DateFlagsTransform ([#552](https://github.com/tinkoff-ai/etna/pull/552))
- Add plot_residuals ([#539](https://github.com/tinkoff-ai/etna/pull/539))
-
- Create `PerSegmentBaseModel`, `PerSegmentPredictionIntervalModel` ([#537](https://github.com/tinkoff-ai/etna/pull/537))
Expand Down
29 changes: 27 additions & 2 deletions etna/transforms/timestamp/date_flags.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -17,6 +17,7 @@ def __init__(
self,
day_number_in_week: Optional[bool] = True,
day_number_in_month: Optional[bool] = True,
day_number_in_year: Optional[bool] = False,
week_number_in_month: Optional[bool] = False,
week_number_in_year: Optional[bool] = False,
month_number_in_year: Optional[bool] = False,
Expand All @@ -34,6 +35,8 @@ def __init__(
if True, add column with weekday info to feature dataframe in transform
day_number_in_month:
if True, add column with day info to feature dataframe in transform
day_number_in_year:
if True, add column with number of day in a year, leap year numeration, starts with 1
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

May you rewrite it to make it clear

week_number_in_month:
if True, add column with week number (in month context) to feature dataframe in transform
week_number_in_year:
Expand Down Expand Up @@ -76,6 +79,7 @@ def __init__(
[
day_number_in_week,
day_number_in_month,
day_number_in_year,
week_number_in_month,
week_number_in_year,
month_number_in_year,
Expand All @@ -87,13 +91,14 @@ def __init__(
):
raise ValueError(
f"{type(self).__name__} feature does nothing with given init args configuration, "
f"at least one of day_number_in_week, day_number_in_month, week_number_in_month, "
f"at least one of day_number_in_week, day_number_in_month, day_number_in_year, week_number_in_month, "
f"week_number_in_year, month_number_in_year, year_number, is_weekend should be True or any of "
f"special_days_in_week, special_days_in_month should be not empty."
)

self.day_number_in_week = day_number_in_week
self.day_number_in_month = day_number_in_month
self.day_number_in_year = day_number_in_year
self.week_number_in_month = week_number_in_month
self.week_number_in_year = week_number_in_year
self.month_number_in_year = month_number_in_year
Expand All @@ -109,6 +114,7 @@ def __init__(
self._empty_parameters = dict(
day_number_in_week=False,
day_number_in_month=False,
day_number_in_year=False,
week_number_in_month=False,
week_number_in_year=False,
month_number_in_year=False,
Expand Down Expand Up @@ -156,6 +162,11 @@ def transform(self, df: pd.DataFrame) -> pd.DataFrame:
timestamp_series=timestamp_series
)

if self.day_number_in_year:
features[self._get_column_name("day_number_in_year")] = self._get_day_number_in_year(
timestamp_series=timestamp_series
)

if self.week_number_in_month:
features[self._get_column_name("week_number_in_month")] = self._get_week_number_in_month(
timestamp_series=timestamp_series
Expand Down Expand Up @@ -228,11 +239,25 @@ def _get_day_number_in_month(timestamp_series: pd.Series) -> np.ndarray:
"""Generate an array with the number of the day in the month."""
return timestamp_series.apply(lambda x: x.day).values

@staticmethod
def _get_day_number_in_year(timestamp_series: pd.Series) -> np.ndarray:
"""Generate an array with the number of the day in a year, lear year numeration, starts with 1."""
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Suggested change
"""Generate an array with the number of the day in a year, lear year numeration, starts with 1."""
"""Generate an array with the numbers of the days in a year, leap year numeration, starts with 1."""

Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Get the right variant from the constructor docstring


def leap_year_number(dt: pd.Timestamp) -> int:
"""Return day number with leap year numeration."""
day_of_year = dt.dayofyear
if not dt.is_leap_year and dt.month >= 3:
return day_of_year + 1
else:
return day_of_year

return timestamp_series.apply(leap_year_number).values

@staticmethod
def _get_week_number_in_month(timestamp_series: pd.Series) -> np.ndarray:
"""Generate an array with the week number in the month."""

def week_of_month(dt: pd.Timestamp) -> float:
def week_of_month(dt: pd.Timestamp) -> int:
"""Return week of month number.

How it works:
Expand Down
17 changes: 14 additions & 3 deletions tests/test_transforms/test_timestamp/test_dateflags_transform.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -17,6 +17,7 @@
INIT_PARAMS_TEMPLATE = {
"day_number_in_week": False,
"day_number_in_month": False,
"day_number_in_year": False,
"week_number_in_year": False,
"week_number_in_month": False,
"month_number_in_year": False,
Expand Down Expand Up @@ -44,6 +45,9 @@ def dateflags_true_df() -> pd.DataFrame:
df = dataframes[i]
df[f"regressor_{out_column}_day_number_in_week"] = df["timestamp"].dt.weekday
df[f"regressor_{out_column}_day_number_in_month"] = df["timestamp"].dt.day
df[f"regressor_{out_column}_day_number_in_year"] = df["timestamp"].apply(
lambda dt: dt.dayofyear + 1 if not dt.is_leap_year and dt.month >= 3 else dt.dayofyear
)
df[f"regressor_{out_column}_week_number_in_year"] = df["timestamp"].dt.week
df[f"regressor_{out_column}_month_number_in_year"] = df["timestamp"].dt.month
df[f"regressor_{out_column}_year_number"] = df["timestamp"].dt.year
Expand Down Expand Up @@ -93,6 +97,7 @@ def test_invalid_arguments_configuration():
_ = DateFlagsTransform(
day_number_in_month=False,
day_number_in_week=False,
day_number_in_year=False,
week_number_in_month=False,
week_number_in_year=False,
month_number_in_year=False,
Expand All @@ -109,6 +114,7 @@ def test_repr():
transform = DateFlagsTransform(
day_number_in_week=True,
day_number_in_month=True,
day_number_in_year=False,
week_number_in_month=False,
week_number_in_year=False,
month_number_in_year=True,
Expand All @@ -119,9 +125,9 @@ def test_repr():
)
transform_repr = transform.__repr__()
true_repr = (
f"{transform_class_repr}(day_number_in_week = True, day_number_in_month = True, week_number_in_month = False, "
f"week_number_in_year = False, month_number_in_year = True, year_number = True, is_weekend = True, special_days_in_week = (1, 2), "
f"special_days_in_month = (12,), out_column = None, )"
f"{transform_class_repr}(day_number_in_week = True, day_number_in_month = True, day_number_in_year = False, "
f"week_number_in_month = False, week_number_in_year = False, month_number_in_year = True, year_number = True, "
f"is_weekend = True, special_days_in_week = (1, 2), special_days_in_month = (12,), out_column = None, )"
)
assert transform_repr == true_repr

Expand All @@ -131,6 +137,7 @@ def test_repr():
(
["day_number_in_week"],
["day_number_in_month"],
["day_number_in_year"],
["week_number_in_year"],
["week_number_in_month"],
["month_number_in_year"],
Expand All @@ -139,6 +146,7 @@ def test_repr():
[
"day_number_in_week",
"day_number_in_month",
"day_number_in_year",
"week_number_in_year",
"week_number_in_month",
"month_number_in_year",
Expand Down Expand Up @@ -173,6 +181,7 @@ def test_interface_correct_args_out_column(true_params: List[str], train_df: pd.
(
["day_number_in_week"],
["day_number_in_month"],
["day_number_in_year"],
["week_number_in_year"],
["week_number_in_month"],
["month_number_in_year"],
Expand All @@ -181,6 +190,7 @@ def test_interface_correct_args_out_column(true_params: List[str], train_df: pd.
[
"day_number_in_week",
"day_number_in_month",
"day_number_in_year",
"week_number_in_year",
"week_number_in_month",
"month_number_in_year",
Expand Down Expand Up @@ -230,6 +240,7 @@ def test_interface_correct_args_repr(true_params: List[str], train_df: pd.DataFr
(
{"day_number_in_week": True},
{"day_number_in_month": True},
{"day_number_in_year": True},
{"week_number_in_year": True},
{"week_number_in_month": True},
{"month_number_in_year": True},
Expand Down