tangent_works.TimeSeries
class tangent_works.TimeSeries(data=pandas.DataFrame, timestamp_column=None, group_key_columns=None)
A class used to represent a time series dataset for Tangent.
Parameters
data : pandas.DataFrame
Dataframe with timeseries data
timestamp_column: str, default None
String indicating the datetime column of the time series.
group_key_columns: list[str], default None
List of strings indicating the categorical columns that will acts as keys for grouped timeseries analysis.
Attributes
timestamps : pd.Series
Series of timestamps in the dataset.
columns : pd.Index
List of column names in the dataset.
sampling_period : SamplingPeriod, optional
Sampling period of the dataset. If not provided, sampling period is detected automatically.
Methods
validate(inplace)
Validates correctness of the dataset (column names and types, timestamps and data).
validate_timestamps_alignment(inplace)
Validates alignment of timestamps with respect to the sampling period.
time_scaling(configuration)
Resamples the dataset according to the provided configuration.
imputation(configuration)
Imputes missing values in the dataset according to the provided imputation configuration.