​

module nixtla.nixtla_client

​

Global Variables

• TYPE_CHECKING

​

class FinetunedModel

​

property model_extra

​

property model_fields_set

​

class AuditDataSeverity

​

class ApiError

​

method __init__

__init__(status_code: Optional[int] = None, body: Optional[Any] = None)

​

class NixtlaClient

​

method __init__

__init__(
    api_key: Optional[str] = None,
    base_url: Optional[str] = None,
    timeout: Optional[int] = 60,
    max_retries: int = 6,
    retry_interval: int = 10,
    max_wait_time: int = 360
)

• api_key (str, optional): The authorization API key to interact with the Nixtla API. If not provided, will use the NIXTLA_API_KEY environment variable.
• base_url (str, optional): Custom base URL. If not provided, will use the NIXTLA_BASE_URL environment variable.
• timeout (int, optional): Request timeout in seconds. Set to None to disable it. Defaults to 60.
• max_retries (int, optional): The maximum number of attempts to make when calling the API before giving up. It defines how many times the client will retry the API call if it fails. Default value is 6, indicating the client will attempt the API call up to 6 times in total. Defaults to 60.
• retry_interval (int, optional): The interval in seconds between consecutive retry attempts. This is the waiting period before the client tries to call the API again after a failed attempt. Default value is 10 seconds, meaning the client waits for 10 seconds between retries. Defaults to 10.
• max_wait_time (int, optional): The maximum total time in seconds that the client will spend on all retry attempts before giving up. This sets an upper limit on the cumulative waiting time for all retry attempts. If this time is exceeded, the client will stop retrying and raise an exception. Default value is 360 seconds, meaning the client will cease retrying if the total time spent on retries exceeds 360 seconds. The client throws a ReadTimeout error after 60 seconds of inactivity. If you want to catch these errors, use max_wait_time >> 60. Defaults to 360.

​

method audit_data

audit_data(
    df: ~AnyDFType,
    freq: Union[str, int, BaseOffset],
    id_col: str = 'unique_id',
    time_col: str = 'ds',
    target_col: str = 'y',
    start: Union[str, int, date, datetime] = 'per_serie',
    end: Union[str, int, date, datetime] = 'global'
) → tuple[bool, dict[str, Union[DataFrame, DataFrame]], dict[str, Union[DataFrame, DataFrame]]]

• df (pandas or polars DataFrame): The dataframe to be audited.
• freq (str, int or pandas offset): Frequency of the timestamps.
• Must be specified. See [pandas' available frequencies](https: //pandas.pydata.org/pandas-docs/stable/user_guide/timeseries.html#offset-aliases).
• id_col (str): Column that identifies each series. Defaults to ‘unique_id’.
• time_col (str): Column that identifies each timestep, its values can be timestamps or integers. Defaults to ‘ds’.
• target_col (str): Column that contains the target. Defaults to ‘y’. start (Union[str, int, datetime.date, datetime.datetime], optional): Initial timestamp for the series. * ‘per_serie’ uses each series first timestamp * ‘global’ uses the first timestamp seen in the data * Can also be a specific timestamp or integer, e.g. ‘2000-01-01’, 2000 or datetime(2000, 1, 1) , by default “per_serie” end (Union[str, int, datetime.date, datetime.datetime], optional): Final timestamp for the series. * ‘per_serie’ uses each series last timestamp * ‘global’ uses the last timestamp seen in the data * Can also be a specific timestamp or integer, e.g. ‘2000-01-01’, 2000 or datetime(2000, 1, 1) , by default “global”

• bool: True if all tests pass, False otherwise
• dict: Dictionary mapping test IDs to error DataFrames for failed tests or None if the test could not be performed.
• dict: Dictionary mapping test IDs to error DataFrames for case-specific tests.

• D001: Test for duplicate rows
• D002: Test for missing dates
• F001: Test for presence of categorical feature columns
• V001: Test for negative values
• V002: Test for leading zeros

​

method clean_data

clean_data(
    df: ~AnyDFType,
    fail_dict: dict[str, Union[DataFrame, DataFrame]],
    case_specific_dict: dict[str, Union[DataFrame, DataFrame]],
    freq: Union[str, int, BaseOffset],
    id_col: str = 'unique_id',
    time_col: str = 'ds',
    target_col: str = 'y',
    clean_case_specific: bool = False,
    agg_dict: Optional[dict[str, Union[str, Callable]]] = None
) → tuple[~AnyDFType, bool, dict[str, Union[DataFrame, DataFrame]], dict[str, Union[DataFrame, DataFrame]]]

• df (AnyDFType): The dataframe to be cleaned
• fail_dict (dict[str, DataFrame]): The failure dictionary from the audit_data method.
• case_specific_dict (dict[str, DataFrame]): The case specific dictionary from the audit_data method.
• freq (str, int or pandas offset): Frequency of the timestamps.
• Must be specified. See [pandas' available frequencies](https: //pandas.pydata.org/pandas-docs/stable/user_guide/timeseries.html#offset-aliases).
• id_col (str): Column that identifies each series. Defaults to ‘unique_id’.
• time_col (str): Column that identifies each timestep, its values can be timestamps or integers. Defaults to ‘ds’.
• target_col (str): Column that contains the target. Defaults to ‘y’.
• clean_case_specific (bool, optional): If True, clean case specific issues. Defaults to False. agg_dict (Optional[dict[str, Union[str, Callable]]], optional): The aggregation methods to use when there are duplicate rows (D001), by default None

• AnyDFType: The cleaned dataframe
• The three outputs from audit_data that are run at the end of cleansing.

• ValueError: Any exceptions during the cleaning process.

​

method cross_validation

cross_validation(
    df: ~AnyDFType,
    h: Annotated[int, Gt(gt=0)],
    freq: Optional[str, int, BaseOffset] = None,
    id_col: str = 'unique_id',
    time_col: str = 'ds',
    target_col: str = 'y',
    level: Optional[list[Union[int, float]]] = None,
    quantiles: Optional[list[float]] = None,
    validate_api_key: bool = False,
    n_windows: Annotated[int, Gt(gt=0)] = 1,
    step_size: Optional[Annotated[int, Gt(gt=0)]] = None,
    finetune_steps: Annotated[int, Ge(ge=0)] = 0,
    finetune_depth: Literal[1, 2, 3, 4, 5] = 1,
    finetune_loss: Literal['default', 'mae', 'mse', 'rmse', 'mape', 'smape'] = 'default',
    finetuned_model_id: Optional[str] = None,
    refit: bool = True,
    clean_ex_first: bool = True,
    hist_exog_list: Optional[list[str]] = None,
    date_features: Union[bool, list[str]] = False,
    date_features_to_one_hot: Union[bool, list[str]] = False,
    model: str = 'timegpt-1',
    num_partitions: Optional[Annotated[int, Gt(gt=0)]] = None
) → ~AnyDFType

• df (pandas or polars DataFrame): The DataFrame on which the function will operate. Expected to contain at least the following columns:
  • time_col: Column name in df that contains the time indices of the time series. This is typically a datetime column with regular intervals, e.g., hourly, daily, monthly data points.
  • target_col: Column name in df that contains the target variable of the time series, i.e., the variable we wish to predict or analyze. Additionally, you can pass multiple time series (stacked in the dataframe) considering an additional column:
  • id_col: Column name in df that identifies unique time series. Each unique value in this column corresponds to a unique time series.
• h (int): Forecast horizon.
• freq (str, int or pandas offset, optional): Frequency of the timestamps. If None, it will be inferred automatically.
• See [pandas' available frequencies](https: //pandas.pydata.org/pandas-docs/stable/user_guide/timeseries.html#offset-aliases). Defaults to None.
• id_col (str): Column that identifies each series. Defaults to ‘unique_id’.
• time_col (str): Column that identifies each timestep, its values can be timestamps or integers. Defaults to ‘ds’.
• target_col (str): Column that contains the target. Defaults to ‘y’.
• level (list[float], optional): Confidence levels between 0 and 100 for prediction intervals. Defaults to None.
• quantiles (list[float], optional): Quantiles to forecast, list between (0, 1). level and quantiles should not be used simultaneously. The output dataframe will have the quantile columns formatted as TimeGPT-q-(100 * q) for each q. 100 * q represents percentiles but we choose this notation to avoid having dots in column names. Defaults to None.
• validate_api_key (bool): If True, validates api_key before sending requests. Defaults to False.
• n_windows (int): Number of windows to evaluate. Defaults to 1.
• step_size (int, optional): Step size between each cross validation window. If None it will be equal to h. Defaults to None.
• finetune_steps (int): Number of steps used to finetune learning TimeGPT in the new data. Defaults to 0.
• finetune_depth (int): The depth of the finetuning. Uses a scale from 1 to 5, where 1 means little finetuning, and 5 means that the entire model is finetuned. Defaults to 1.
• finetune_loss (str): Loss function to use for finetuning. Options
• are: default, mae, mse, rmse, mape, and smape. Defaults to ‘default’.
• finetuned_model_id (str, optional): ID of previously fine-tuned model to use. Defaults to None.
• finetuned_model_id (str, optional): ID of previously fine-tuned model to use. Defaults to None. refit (bool): Fine-tune the model in each window. If False, only fine-tunes on the first window. Only used if finetune_steps > 0. Defaults to True. clean_ex_first (bool): Clean exogenous signal before making forecasts using TimeGPT. Defaults to True. hist_exog_list (list[str], optional): Column names of the historical exogenous features. Defaults to None.
• date_features (bool or list[str] or callable, optional): Features computed from the dates. Can be pandas date attributes or functions that will take the dates as input. If True automatically adds most used date features for the frequency of df. Defaults to False.
• date_features_to_one_hot (bool or list[str]): Apply one-hot encoding to these date features. If date_features=True, then all date features are one-hot encoded by default. Defaults to False.
• model (str): Model to use as a string. Options are: timegpt-1, and timegpt-1-long-horizon. We recommend using timegpt-1-long-horizon for forecasting if you want to predict more than one seasonal period given the frequency of your data. Defaults to ‘timegpt-1’. num_partitions (int): Number of partitions to use. If None, the number of partitions will be equal to the available parallel resources in distributed environments. Defaults to None.

​

method delete_finetuned_model

delete_finetuned_model(finetuned_model_id: str) → bool

• finetuned_model_id (str): ID of the fine-tuned model to be deleted.

• bool: Whether delete was successful.

​

method detect_anomalies

detect_anomalies(
    df: ~AnyDFType,
    freq: Optional[str, int, BaseOffset] = None,
    id_col: str = 'unique_id',
    time_col: str = 'ds',
    target_col: str = 'y',
    level: Union[int, float] = 99,
    finetuned_model_id: Optional[str] = None,
    clean_ex_first: bool = True,
    validate_api_key: bool = False,
    date_features: Union[bool, list[str]] = False,
    date_features_to_one_hot: Union[bool, list[str]] = False,
    model: str = 'timegpt-1',
    num_partitions: Optional[Annotated[int, Gt(gt=0)]] = None
) → ~AnyDFType

• df (pandas or polars DataFrame): The DataFrame on which the function will operate. Expected to contain at least the following columns:
  • time_col: Column name in df that contains the time indices of the time series. This is typically a datetime column with regular intervals, e.g., hourly, daily, monthly data points.
  • target_col: Column name in df that contains the target variable of the time series, i.e., the variable we wish to predict or analyze. Additionally, you can pass multiple time series (stacked in the dataframe) considering an additional column:
  • id_col: Column name in df that identifies unique time series. Each unique value in this column corresponds to a unique time series.
• freq (str, int, pandas offset, optional): Frequency of the timestamps. If None, it will be inferred automatically.
• See [pandas' available frequencies](https: //pandas.pydata.org/pandas-docs/stable/user_guide/timeseries.html#offset-aliases). Defaults to None.
• id_col (str): Column that identifies each series. Defaults to ‘unique_id’.
• time_col (str): Column that identifies each timestep, its values can be timestamps or integers. Defaults to ‘ds’.
• target_col (str): Column that contains the target. Defaults to ‘y’.
• level (float): Confidence level between 0 and 100 for detecting the anomalies. Defaults to 99.
• finetuned_model_id (str, optional): ID of previously fine-tuned model to use. Defaults to None.
• clean_ex_first (bool): Clean exogenous signal before making forecasts using TimeGPT. Defaults to True. validate_api_key (bool): If True, validates api_key before sending requests. Defaults to False.
• date_features (bool or list[str] or callable, optional): Features computed from the dates. Can be pandas date attributes or functions that will take the dates as input. If True automatically adds most used date features for the frequency of df. Defaults to False.
• date_features_to_one_hot (bool or list[str]): Apply one-hot encoding to these date features. If date_features=True, then all date features are one-hot encoded by default. Defaults to False.
• model (str): str (default=‘timegpt-1’)
• Model to use as a string. Options are: timegpt-1, and timegpt-1-long-horizon. We recommend using timegpt-1-long-horizon for forecasting if you want to predict more than one seasonal period given the frequency of your data. Defaults to ‘timegpt-1’.
• num_partitions (int): Number of partitions to use. If None, the number of partitions will be equal to the available parallel resources in distributed environments. Defaults to None.

​

method detect_anomalies_online

detect_anomalies_online(
    df: ~AnyDFType,
    h: Annotated[int, Gt(gt=0)],
    detection_size: Annotated[int, Gt(gt=0)],
    threshold_method: Literal['univariate', 'multivariate'] = 'univariate',
    freq: Optional[str, int, BaseOffset] = None,
    id_col: str = 'unique_id',
    time_col: str = 'ds',
    target_col: str = 'y',
    level: Union[int, float] = 99,
    clean_ex_first: bool = True,
    step_size: Optional[Annotated[int, Gt(gt=0)]] = None,
    finetune_steps: Annotated[int, Ge(ge=0)] = 0,
    finetune_depth: Literal[1, 2, 3, 4, 5] = 1,
    finetune_loss: Literal['default', 'mae', 'mse', 'rmse', 'mape', 'smape'] = 'default',
    hist_exog_list: Optional[list[str]] = None,
    date_features: Union[bool, list[str]] = False,
    date_features_to_one_hot: Union[bool, list[str]] = False,
    model: str = 'timegpt-1',
    refit: bool = False,
    num_partitions: Optional[Annotated[int, Gt(gt=0)]] = None
) → ~AnyDFType

• time_col: Column name in df that contains the time indices of the time series. This is typically a datetime column with regular intervals, e.g., hourly, daily, monthly data points.
• target_col: Column name in df that contains the target variable of the time series, i.e., the variable we wish to predict or analyze.
• id_col: Column name in df that identifies unique time series. Each unique value in this column corresponds to a unique time series.
• h (int): Forecast horizon.
• detection_size (int): The length of the sequence where anomalies will be detected starting from the end of the dataset.
• threshold_method (str, optional): The method used to calculate the intervals for anomaly detection. Use univariate to flag anomalies independently for each series in the dataset. Use multivariate to have a global threshold across all series in the dataset. For this method, all series must have the same length. Defaults to ‘univariate’.
• freq (str, optional): Frequency of the data. By default, the freq
• will be inferred automatically. See [pandas' available frequencies](https: //pandas.pydata.org/pandas-docs/stable/user_guide/timeseries.html#offset-aliases).
• id_col (str, optional): Column that identifies each series. Defaults to ‘unique_id’
• time_col (str, optional): Column that identifies each timestep, its values can be timestamps or integers. Defaults to ‘ds’.
• target_col (str, optional): Column that contains the target. Defaults to ‘y’. level (float, optional): Confidence level between 0 and 100 for detecting the anomalies. Defaults to 99.
• clean_ex_first (bool, optional): Clean exogenous signal before making forecasts using TimeGPT. Defaults to True.
• step_size (int, optional): Step size between each cross validation window. If None it will be equal to h. Defaults to None.
• finetune_steps (int): Number of steps used to finetune TimeGPT in the new data. Defaults to 0.
• finetune_depth (int): The depth of the finetuning. Uses a scale from 1 to 5, where 1 means little finetuning, and 5 means that the entire model is finetuned. Defaults to 1.
• finetune_loss (str): Loss function to use for finetuning.
• Options are: default, mae, mse, rmse, mape, and smape. Defaults to ‘default’.
• hist_exog_list (list[str], optional): Column names of the historical exogenous features. Defaults to None.
• date_features (bool or list[str] or callable, optional): Features computed from the dates. Can be pandas date attributes or functions that will take the dates as input. If True automatically adds most used date features for the frequency of df. Defaults to False.
• date_features_to_one_hot (bool or list[str]): Apply one-hot encoding to these date features. If date_features=True, then all date features are one-hot encoded by default. Defaults to False.
• model (str, optional): Model to use as a string. Options are: timegpt-1, and timegpt-1-long-horizon. We recommend using timegpt-1-long-horizon for forecasting if you want to predict more than one seasonal period given the frequency of your data. Defaults to ‘timegpt-1’.
• refit (bool, optional): Fine-tune the model in each window. If False, only fine-tunes on the first window. Only used if finetune_steps > 0. Defaults to False. num_partitions (int): Number of partitions to use. If None, the number of partitions will be equal to the available parallel resources in distributed environments. Defaults to None.

​

method finetune

finetune(
    df: Union[DataFrame, DataFrame],
    freq: Optional[str, int, BaseOffset] = None,
    id_col: str = 'unique_id',
    time_col: str = 'ds',
    target_col: str = 'y',
    finetune_steps: Annotated[int, Ge(ge=0)] = 10,
    finetune_depth: Literal[1, 2, 3, 4, 5] = 1,
    finetune_loss: Literal['default', 'mae', 'mse', 'rmse', 'mape', 'smape'] = 'default',
    output_model_id: Optional[str] = None,
    finetuned_model_id: Optional[str] = None,
    model: str = 'timegpt-1'
) → str

• df (pandas or polars DataFrame): The DataFrame on which the function will operate. Expected to contain at least the following columns:
  • time_col: Column name in df that contains the time indices of the time series. This is typically a datetime column with regular intervals, e.g., hourly, daily, monthly data points.
  • target_col: Column name in df that contains the target variable of the time series, i.e., the variable we wish to predict or analyze. Additionally, you can pass multiple time series (stacked in the dataframe) considering an additional column:
  • id_col: Column name in df that identifies unique time series. Each unique value in this column corresponds to a unique time series.
• freq (str, int, pandas offset, optional): Frequency of the timestamps. If None, it will be inferred automatically.
• See [pandas' available frequencies](https: //pandas.pydata.org/pandas-docs/stable/user_guide/timeseries.html#offset-aliases). Defaults to None.
• id_col (str): Column that identifies each series. Defaults to ‘unique_id’.
• time_col (str): Column that identifies each timestep, its values can be timestamps or integers. Defaults to ‘ds’.
• target_col (str): Column that contains the target. Defaults to ‘y’.
• finetune_steps (int): Number of steps used to finetune learning TimeGPT in the new data. Defaults to 10.
• finetune_depth (int): The depth of the finetuning. Uses a scale from 1 to 5, where 1 means little finetuning, and 5 means that the entire model is finetuned. Defaults to 1.
• finetune_loss (str): Loss function to use for finetuning. Options
• are: default, mae, mse, rmse, mape, and smape. Defaults to ‘default’.
• output_model_id (str, optional): ID to assign to the fine-tuned model. If None, an UUID is used. Defaults to None.
• finetuned_model_id (str, optional): ID of previously fine-tuned model to use as base. Defaults to None. model (str):
• Model to use as a string. Options are: timegpt-1, and timegpt-1-long-horizon. We recommend using timegpt-1-long-horizon for forecasting if you want to predict more than one seasonal period given the frequency of your data. Defaults to ‘timegpt-1’.

• str: ID of the fine-tuned model

​

method finetuned_model

finetuned_model(finetuned_model_id: str) → FinetunedModel

• finetuned_model_id (str): ID of the fine-tuned model to get metadata from.

• FinetunedModel: Fine-tuned model metadata.

​

method finetuned_models

finetuned_models(as_df: bool = False) → Union[list[FinetunedModel], DataFrame]

• as_df (bool): Return the fine-tuned models as a pandas dataframe.

• List of FinetunedModel: List of available fine-tuned models.

​

method forecast

forecast(
    df: ~AnyDFType,
    h: Annotated[int, Gt(gt=0)],
    freq: Optional[str, int, BaseOffset] = None,
    id_col: str = 'unique_id',
    time_col: str = 'ds',
    target_col: str = 'y',
    X_df: Optional[~AnyDFType] = None,
    level: Optional[list[Union[int, float]]] = None,
    quantiles: Optional[list[float]] = None,
    finetune_steps: Annotated[int, Ge(ge=0)] = 0,
    finetune_depth: Literal[1, 2, 3, 4, 5] = 1,
    finetune_loss: Literal['default', 'mae', 'mse', 'rmse', 'mape', 'smape'] = 'default',
    finetuned_model_id: Optional[str] = None,
    clean_ex_first: bool = True,
    hist_exog_list: Optional[list[str]] = None,
    validate_api_key: bool = False,
    add_history: bool = False,
    date_features: Union[bool, list[Union[str, Callable]]] = False,
    date_features_to_one_hot: Union[bool, list[str]] = False,
    model: str = 'timegpt-1',
    num_partitions: Optional[Annotated[int, Gt(gt=0)]] = None,
    feature_contributions: bool = False
) → ~AnyDFType

• df (pandas or polars DataFrame): The DataFrame on which the function will operate. Expected to contain at least the following columns:
  • time_col: Column name in df that contains the time indices of the time series. This is typically a datetime column with regular intervals, e.g., hourly, daily, monthly data points.
  • target_col: Column name in df that contains the target variable of the time series, i.e., the variable we wish to predict or analyze. Additionally, you can pass multiple time series (stacked in the dataframe) considering an additional column:
  • id_col: Column name in df that identifies unique time series. Each unique value in this column corresponds to a unique time series.
• h (int): Forecast horizon.
• freq (str, int or pandas offset, optional): Frequency of the timestamps. If None, it will be inferred automatically.
• See [pandas' available frequencies](https: //pandas.pydata.org/pandas-docs/stable/user_guide/timeseries.html#offset-aliases). Defaults to None.
• id_col (str): Column that identifies each series. Defaults to ‘unique_id’.
• time_col (str): Column that identifies each timestep, its values can be timestamps or integers. Defaults to ‘ds’.
• target_col (str): Column that contains the target. Defaults to ‘y’. X_df (pandas or polars DataFrame, optional): DataFrame with [unique_id, ds] columns and df’s future exogenous. Defaults to None.
• level (list[float], optional): Confidence levels between 0 and 100 for prediction intervals. Defaults to None.
• quantiles (list[float], optional): Quantiles to forecast, list between (0, 1). level and quantiles should not be used simultaneously. The output dataframe will have the quantile columns formatted as TimeGPT-q-(100 * q) for each q. 100 * q represents percentiles but we choose this notation to avoid having dots in column names. Defaults to None.
• finetune_steps (int): Number of steps used to finetune learning TimeGPT in the new data. Defaults to 0.
• finetune_depth (int): The depth of the finetuning. Uses a scale from 1 to 5, where 1 means little finetuning, and 5 means that the entire model is finetuned. Defaults to 1.
• finetune_loss (str): Loss function to use for finetuning. Options
• are: default, mae, mse, rmse, mape, and smape. Defaults to ‘default’.
• finetuned_model_id (str, optional): ID of previously fine-tuned model to use. Defaults to None.
• clean_ex_first (bool): Clean exogenous signal before making forecasts using TimeGPT. Defaults to True.
• hist_exog_list (list[str], optional): Column names of the historical exogenous features. Defaults to None. validate_api_key (bool): If True, validates api_key before sending requests. Defaults to False.
• add_history (bool): Return fitted values of the model. Defaults to False.
• date_features (bool or list[str] or callable, optional): Features computed from the dates. Can be pandas date attributes or functions that will take the dates as input. If True automatically adds most used date features for the frequency of df. Defaults to False.
• date_features_to_one_hot (bool or list[str]): Apply one-hot encoding to these date features. If date_features=True, then all date features are one-hot encoded by default. Defaults to False.
• model (str): Model to use as a string. Options are: timegpt-1, and timegpt-1-long-horizon. We recommend using timegpt-1-long-horizon for forecasting if you want to predict more than one seasonal period given the frequency of your data. Defaults to ‘timegpt-1’. num_partitions (int): Number of partitions to use. If None, the number of partitions will be equal to the available parallel resources in distributed environments. Defaults to None.
• feature_contributions (bool): Compute SHAP values. Gives access to computed SHAP values to explain the impact of features on the final predictions. Defaults to False.

​

method plot

plot(
    df: Optional[DataFrame, DataFrame] = None,
    forecasts_df: Optional[DataFrame, DataFrame] = None,
    id_col: str = 'unique_id',
    time_col: str = 'ds',
    target_col: str = 'y',
    unique_ids: Union[list[str], NoneType, ndarray] = None,
    plot_random: bool = True,
    max_ids: int = 8,
    models: Optional[list[str]] = None,
    level: Optional[list[Union[int, float]]] = None,
    max_insample_length: Optional[int] = None,
    plot_anomalies: bool = False,
    engine: Literal['matplotlib', 'plotly', 'plotly-resampler'] = 'matplotlib',
    resampler_kwargs: Optional[dict] = None,
    ax: Optional[ForwardRef('Axes'), ndarray, ForwardRef('Figure')] = None
)

• df (pandas or polars DataFrame): The DataFrame on which the function will operate. Expected to contain at least the following columns:
  • time_col: Column name in df that contains the time indices of the time series. This is typically a datetime column with regular intervals, e.g., hourly, daily, monthly data points.
  • target_col: Column name in df that contains the target variable of the time series, i.e., the variable we wish to predict or analyze. Additionally, you can pass multiple time series (stacked in the dataframe) considering an additional column:
  • id_col: Column name in df that identifies unique time series. Each unique value in this column corresponds to a unique time series.
• forecasts_df (pandas or polars DataFrame, optional): DataFrame with columns [unique_id, ds] and models. Defaults to None.
• id_col (str): Column that identifies each series. Defaults to ‘unique_id’.
• time_col (str): Column that identifies each timestep, its values can be timestamps or integers. Defaults to ‘ds’.
• target_col (str): Column that contains the target. Defaults to ‘y’.
• unique_ids (list[str], optional): Time Series to plot. If None, time series are selected randomly. Defaults to None. plot_random (bool): Select time series to plot randomly. Defaults to True. max_ids (int): Maximum number of ids to plot. Defaults to 8.
• models (list[str], optional): List of models to plot. Defaults to None.
• level (list[float], optional): List of prediction intervals to plot if paseed. Defaults to None.
• max_insample_length (int, optional): Max number of train/insample observations to be plotted. Defaults to None.
• plot_anomalies (bool): Plot anomalies for each prediction interval. Defaults to False.
• engine (str): Library used to plot. ‘matplotlib’, ‘plotly’ or ‘plotly-resampler’. Defaults to ‘matplotlib’.
• resampler_kwargs (dict): Kwargs to be passed to plotly-resampler constructor. For further custumization (“show_dash”) call the method, store the plotting object and add the extra arguments to its show_dash method. ax (matplotlib axes, array of matplotlib axes or plotly Figure,
• optional): Object where plots will be added. Defaults to None.

​

method usage

usage() → dict[str, dict[str, int]]

• dict: Consumed requests and limits by minute and month.

​

method validate_api_key

validate_api_key(log: bool = True) → bool

• log (bool): Show the endpoint’s response. Defaults to True.

• bool: Whether API key is valid.