paddlets.analysis.time_domain

class Seasonality(period: Union[None, int] = None, nlags: int = 300, alpha: float = 0.05, mode: str = 'additive', order: int = 1, **kwargs)[source]

Bases: Analyzer

Compute the seasonality period of given columns.

Parameters

period (int) – The period of the data. If None(by default), we will calculate a unique seasonality period.
nlags (int) – Number of lags to return autocorrelation for, default=300.
alpha (float) – The confidence intervals for the seasonality. default=0.05.
mode (str) – Type of seasonal component. Abbreviations are accepted. Optional(“additive”, “multiplicative”).
order (int) – How many points on each side to use for the comparisont o consider comparator(n, n+x) to be True.
kwargs – Other parameters.

analyze(X: Union[Series, DataFrame]) → Union[Any, Series][source]

Compute the seasonality period of given columns

Parameters: X (pd.Series|pd.DataFrame) – columns to be analyzed
Returns: The seasonality period and seasonality values
Return type: (dict, dict)
Raises: ValueError –

plot() → pyplot[source]

display seasonality result.

Parameters: None –
Returns: The seasonality figure
Return type: plt(matplotlib.pyplot object)
Raises: None –

classmethod get_properties() → Dict[source]

Get the properties of the analyzer.

Returns: Dict

class Acf(nlags: int = 300, alpha: float = 0.05, **kwargs)[source]

Bases: Analyzer

Compute the acf values of given columns.

Parameters

nlags (int) – Number of lags to return autocorrelation for, default=300.
alpha (float) – The confidence intervals for the acf. default=0.05.
kwargs – Other parameters.

analyze(X: Union[Series, DataFrame]) → Union[Any, Series][source]

Compute the acf values of given columns

Parameters: X (pd.Series|pd.DataFrame) – columns to be analyzed
Returns: The acf values and confident values
Return type: dict
Raises: ValueError –

plot() → pyplot[source]

display acf result.

Parameters: None –
Returns: The acf figure
Return type: plt(matplotlib.pyplot object)
Raises: None –

classmethod get_properties() → Dict[source]

Get the properties of the analyzer.

Returns: Dict

class Correlation(method: str = 'pearson', lag: int = 0, lag_cols: Optional[Union[str, List[str]]] = [], **kwargs)[source]

Bases: Analyzer

Compute the correlation values of given columns.

Parameters

method (str) – {‘pearson’, ‘kendall’, ‘spearman’} or callable
lag (int) – lag time points.
lag_cols (List[str], str) – columns that need lag.
kwargs – Other parameters.

analyze(X: DataFrame) → Union[Any, Series][source]

Compute the correlation values of given columns.

Parameters: X (pd.DataFrame) – columns to be analyzed
Returns: The acf values and confident values
Return type: dict
Raises: ValueError –

plot() → pyplot[source]

display correlation result.

Parameters: None –
Returns: The correlation figure
Return type: plt(matplotlib.pyplot object)
Raises: None –

classmethod get_properties() → Dict[source]

Get the properties of the analyzer.

Returns: Dict