Add option day_number_in_year to DateFlagsTransform

CHANGELOG.md

@@ -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))

etna/transforms/timestamp/date_flags.py

@@ -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,
@@ -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 with leap year numeration (values from 1 to 366)
         week_number_in_month:
             if True, add column with week number (in month context) to feature dataframe in transform
         week_number_in_year:
@@ -52,8 +55,8 @@ def __init__(
             with flag that shows given date is a special day
         out_column:
             base for the name of created columns;
-            if set the final name is '{out_column}_{feature_name}', don't forget to add 'regressor_' prefix;
-            if don't set, name will be 'regressor_{transform.__repr__()}',
+            if set the final name is '{out_column}_{feature_name}';
+            if don't set, name will be `transform.__repr__()`,
             repr will be made for transform that creates exactly this column
 
         Notes
@@ -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,
@@ -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
@@ -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,
@@ -123,7 +129,7 @@ def _get_column_name(self, feature_name: str) -> str:
             init_parameters = deepcopy(self._empty_parameters)
             init_parameters[feature_name] = self.__dict__[feature_name]
             temp_transform = DateFlagsTransform(**init_parameters, out_column=self.out_column)  # type: ignore
-            return f"regressor_{temp_transform.__repr__()}"
+            return temp_transform.__repr__()
         else:
             return f"{self.out_column}_{feature_name}"
 
@@ -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
@@ -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 number of day in a year with leap year numeration (values from 1 to 366)."""
+
+        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: