_OneSegmentSpecialDaysTransform¶

class _OneSegmentSpecialDaysTransform(find_special_weekday: bool = True, find_special_month_day: bool = True)[source]¶

Search for anomalies in values, marked this days as 1 (and return new column with 1 in corresponding places).

Notes

You can read more about other anomalies detection methods in: Time Series of Price Anomaly Detection

Create instance of _OneSegmentSpecialDaysTransform.

Parameters

find_special_weekday (bool) – flag, if True, find special weekdays in transform
find_special_month_day (bool) – flag, if True, find special monthdays in transform

Raises

ValueError: – if all the modes are False

Inherited-members

Methods

`fit`(df)	Fit _OneSegmentSpecialDaysTransform with data from df.
`fit_transform`(df)	Fit and transform Dataframe.
`inverse_transform`(df)	Inverse transform Dataframe.
`set_params`(**params)	Return new object instance with modified parameters.
`to_dict`()	Collect all information about etna object in dict.
`transform`(df)	Transform data from df with _OneSegmentSpecialDaysTransform and generate a column of special day flags.

Fit _OneSegmentSpecialDaysTransform with data from df.

Parameters: df (pd.DataFrame) – value series with index column in timestamp format
Return type: etna.transforms.timestamp.special_days._OneSegmentSpecialDaysTransform

inverse_transform(df: pandas.core.frame.DataFrame) → pandas.core.frame.DataFrame[source]¶

Inverse transform Dataframe.

transform(df: pandas.core.frame.DataFrame) → pandas.core.frame.DataFrame[source]¶

Transform data from df with _OneSegmentSpecialDaysTransform and generate a column of special day flags.

Parameters: df (pd.DataFrame) – value series with index column in timestamp format
Returns: pd.DataFrame with ‘anomaly_weekday’, ‘anomaly_monthday’ or both of them columns no-timestamp indexed that contains 1 at i-th position if i-th day is a special day
Return type: pandas.core.frame.DataFrame